FloMax Software for Cytometry. Operating Manual - Data Analysis. Version 2.3, FloMax Operating Manual Data Analysis 1

Transcription

1 FloMax Software for Cytometry Operating Manual - Data Analysis Version 2.3, FloMax Operating Manual Data Analysis 1

2 Contents General Information...4 Conventions Used in this Manual...5 Introduction to FloMax Software...6 Software Installation...8 Software Registration...9 Screen Elements...10 Introduction One-Parameter Histograms...11 Introduction - Correlated 2 Parameter Dotplots...13 Introduction FloMax Start and Exit...15 Introduction Typical Steps for a Immunophenotype Analysis...16 Introduction Typical Steps for a Kinetic Analysis...17 Introduction Typical Steps for DNA Cell Cycle Analysis...18 Introduction Typical Steps for Microorganism Analysis...19 Opening a Flow Cytometry Data File...20 Saving an Analyzed Flow Cytometric File...21 Printing the Histogram Page...22 Printer Setup...23 Selecting the Page Layout...24 Selecting Histogram Properties...25 Histogram Properties...26 Analysing Channel Contents...28 Absolute Cell Counting / Analysing the Particle Concentration...29 Rerun From File Function...30 Zoom Function...31 Gating - Defining Regions...32 Gating - Applying Gates...33 Gating - Logical Gates...34 Gating - Saving and Loading Gates and Regions...35 Gating - Moving, Resizing, and Deleting Regions...36 Gating - Obtaining Statistics...37 Gating Saving Region Statistics to a File...38 Gating Exporting a Gated File...39 Gating - Color Gating...40 Crosstalk Compensation - Overview...42 Crosstalk Compensation - Setup...43 Crosstalk Compensation Saving and Loading...44 Crosstalk Compensation Exporting a Compensated File...45 Crosstalk Compensation - Notes...46 Cell Cycle Analysis...48 Peak Analysis Numerical Fit Method...49 Peak Analysis Find Peaks Range Method & Batch Analysis...50 Peak Analysis Histogram Export...51 Ratio Analysis...52 Calculated Parameters...53 Calculated Parameters Examples...54 Calculated Parameters Examples...55 Calculated Parameters Function Overview...56 Panels - Overview...57 Panels- Definition...58 Panels - Optimizing...59 Generating Reports Copy & Paste...60 Generating Reports...61 Appendix Software Support...62 Appendix - Mathematical Formulas...63 Appendix Computer Recommendations...74 Appendix Software Specifications...75 Appendix Menu Function Overview...77 Index...78 FloMax Operating Manual Data Analysis 2

3 FloMax Software for Cytometry Operating Manual Data Analysis Version by Partec GmbH, Münster, Germany FloMax Operating Manual Data Analysis 3

4 General Information Questions Disclaimer Trademarks and Copyrights FloMax is made for you! New software versions and you profit from your requests for new features or software changes. Partec is continuously working on software to fulfill your demands. If you have questions concerning this manual or the software, if you find problems associated with FloMax, or if you have a good suggestion to be included in a new version, please let us know by sending an or a note to Partec. FloMax software and this manual are intended to be used by experienced flow cytometrists. FloMax was developed and tested to make flow cytometry data analysis more comfortable. Complex software like FloMax and this manual however do not claim to be completely error-free. As with all software, results obtained with FloMax should be checked and verified critically by the user. Partec recommends to test FloMax for a specific application before running large scale sample series. Partec provides FloMax as is. Partec does not take responsibility that FloMax is suited for a specific application. Partec also does not take responsibility for any direct or indirect damage that could be caused by FloMax or based on results obtained by FloMax. Especially no responsibility can be claimed for data or reagent loss due to the use of FloMax. This manual contains references to names and products from Partec and other companies which are registered trademarks or protected by copyright. Partec GmbH: PAS, PA, CCA, Robby. Quantum Analysis GmbH: FloMax. Cytecs GmbH: CyFlow. Microsoft Corp.: Windows, Word, Excel, PowerPoint, Paint. Hewlett Packard : Deskjet, Laserjet. FloMax Operating Manual Data Analysis 4

5 Conventions Used in this Manual Selection of Menu Functions The location of menu selections is frequently noted in a short form as in the following example: Analysis Calculate Parameters Formula denotes the following menu selection: Notes and Tips Notes and tips are denoted by this symbol. Mouse Operation click single click with the left mouse key. doubleclick two consecutive clicks with the left mouse key. right click single click with the right mouse key. FloMax Operating Manual Data Analysis 5

6 Introduction to FloMax Software What is Partec FloMax? topics are covered by this manual? other manuals are available? if my cytometer isn t a Partec? are the applications for which FloMax can be used? Partec FloMax software is an all-in-one solution for acquisition and analysis of flow cytometric data. FloMax data analysis works with data from flow cytometers that support the flow cytometry file standard (FCS). FloMax operates on computers with Windows 95, 98, and Data acquisition and instrument control by FloMax is provided for Partec Flow Cytometers. FloMax Data Analysis (this manual) covers all aspects concerning the flow cytometric data analysis. Most of the functions are available online during acquisition as well as offline for files already stored. FloMax - Instrument Control and Data Acquisition covers additional functions for use on a Partec flow cytometer. Please consult the Instrument Operating Manuals for your particular flow cytometer for details on the instrument operation. Application Notes and Tutorials are available to get started and contain hints to achieve the best results. FloMax supports the FCS (flow cytometry standard) data format. This makes FloMax suitable for data analysis from any flow cytometer, offering unique functionality and automation. FloMax offers automation for routine use and flexibility for research use for practically any flow cytometric application. The applications cover: Routine Immunophenotype Blood Cell Analysis Leukocyte Counting Rare Event Analysis DNA Cell Cycle Analysis Ploidy Analysis Microorganism Analysis (live/dead) Fermentation Control Particle Concentration Analysis True Volumetric Absolute Counting Particle Size and Fluorescence Distribution Analysis FloMax Operating Manual Data Analysis 6

7 Introduction to FloMax Software What... functions are offered by FloMax? FCS Data Analysis and Report Generation Flexible data display on pages with 1-8 oneand two-parameter histograms and dotplots Multiparameter Gating with logical gates MultiColor Gating Colour Crosstalk Compensation by Software DNA Cell Cycle and Peak Analysis Calculated Parameters Panel System Automated Report Generation in conjunction with MS Word and Excel. additional features offers FloMax in conjunction with Partec flow cytometers and sample automates? Instrument control and data acquisition functions are covered in a separate manual. Instrument Control and Data Acquisition Flow cytometer instrument control Realtime Acquisition True Volumetric Absolute Counting Complete walk-away sample preparation and analysis... should I know before operating FloMax?... can I do in case I have questions or problems? This manual assumes you have basic knowledge about flow cytometry. In the best case a well experienced "flower" is around - so let her/him help you. Many basic books are available about flow cytometry which may help you as well. FloMax makes operation as simple as possible. When you are familiar with Windows and other Windows software, you will find it quite easy to operate FloMax. We at Partec are proud if we may assist you with any question you have with your application. Don t hesitate to contact us! Partec continuously publishes application notes which might contain important hints not covered by this manual. Please inquire for the latest list of application notes. Partec is available for your support. Don t hesitate to contact us by , fax or telephone. FloMax Operating Manual Data Analysis 7

8 Software Installation Shortguide for FloMax Installation 1. Archivate existing FloMax folder: E. g. rename C:\FloMax to C:\FloMax20. IMPORTANT: If you don t archivate your existing version it will be overwritten. FloMax Installation CD. If automatic launching of installation programs is disabled on your computer, the installation program does not start automatically. In this case, select Start-Run in the Windows toolbar and type D:\setup.exe. Press OK and proceed with step Insert the FloMax Installation CD after some seconds, the FloMax installation programm starts. 3. Follow the instructions given by the installation program. Use the defaults whenever possible. - FloMax will be completely installed. - The FloMax icon appears on the desktop (and under Start - Programs). FloMax Operating Manual Data Analysis 8

9 Software Registration Software Registration In order to offer you the best possible support for FloMax, you should register for a full license. A full license allows you to use FloMax on one or more computers in your laboratory which are under your responsibility. Without software registration, FloMax can be run in a demo mode with limited options. About FloMax box. How to get the registration code If you purchased FloMax or a flow cytometer together with FloMax, in order to receive your personal registration code: 1. Make sure FloMax is installed on your computer and start FloMax. - When FloMax is not registered yet, the Software Registration dialog appears asking you for registration. or Select Help-About FloMax in the menu - the About FloMax box appears. Click Register - the Registration dialog appears. Software Registration Dialog. 3. Click Get Registration Code the Software Registration Form appears. 4. Fill out the form completely with your name, address, and all other information. 5. Click Print Registration Form, print the form on a printer. Fax the form to Partec to the fax no. given below. or Call Partec at the phone no. below. or Click Editor and cut and paste the registration form to an . Send the to the address below. This is the preferred way. - Within a short perioud, you will receive the FloMax registration code by Partec. Software Registration Form. Each computer requires a separate registration. For each laboratory computer, a separate registration form should be submitted to Partec. With a full license, you may inquire registration codes for one or more computers in your laboratory you are using. FloMax Registration Partec GmbH, Münster, Germany Phone Fax [email protected] FloMax Operating Manual Data Analysis 9

10 Screen Elements Menu Bar Title Bar Plot Property Box Dialog Selection Frame with Trackers Plot Selection Marks Plots Results Field Buttons Mouse Cursor Status Bar Toolbar Tooltip Scrollbar Elements on the FloMax software screen. FloMax Operating Manual Data Analysis 10

11 Introduction One-Parameter Histograms What is... a parameter?... a one-parameter histogram?... a histogram channel?... the count in a histogram?... a peak?... background in a histogram?... the lower level (L-L)? In flow cytometry, parameter denotes a measured property of the particles. Frequently, a parameter is synonymeous to an optical channel. E.g. an instrument with 6 parameters is equipped with 6 optical detectors. Parameter no. 6 could e.g. be the blue fluorescence parameter that could be called FL4 and mainly be used for DNA analysis. A one- parameter histogram displays the distribution of cell contents, e.g. how many cells contain a given quantity of DNA or bind a given number of antibody molecules. The cell content is assigned to one of many (e.g ) quantity classes or channels. In a oneparameter histogram the channels are represented on the x-axis. The number of cells being assigned to a given channel is referred to as channel content or simply count. In a one-parameter histogram, the count is shown on the y-axis. All cells having about equal quantity of the cell content, e.g. DNA, form a peak. In the case a of typical DNA histogram e.g., one peak represents the G1 and another (with twice the channel value) represents the G2/M phase of the cell cycle. In the case of immunolabelled cells, there is usually one peak for unlabelled (negative) and one peak for labelled (positive) cells. Peaks can be analyzed by drawing ranges or by numerical fits, e.g. in order to know the mean intensity or number of cells in a peak. Histograms sometimes show undesired signals in the lower channels, frequently called noise or background. These signals can originate from cell fragments or other particles resulting from sample preparation. In case of high gains, background can also be caused by not sufficiently clean sheath fluid or by background light. The lower level (L-L) threshold is a means to suppress background signals. Signals below the lower level are rejected from the histogram during acquisition. To exclude noise from a histogram already acquired, a range-gate can be used. FloMax Operating Manual Data Analysis 11

12 Introduction One-Parameter Histograms peak of normal G1-cells counts peak of tumor G1-cells lower level threshold peak of tumor (G2+M)-cells background channels One-parameter histogram with 512 channels. EAT cells (DAPI-staining, HBO-lamp excitation). Background due to cell fragments. One parameter histogram of immunolabelled lymphocytes (green fluorescence excited by Ar-laser). The unlabelled cells ( neg. ) appear in the left peak (autofluorescence and unspecific antibody binding); brightly labelled cells in the right peak. Dimly labelled cells appear between the peaks. Note the log-scale typically used for this kind of measurements. FloMax Operating Manual Data Analysis 12

13 Introduction - Correlated 2 Parameter Dotplots What is a correlated 2 parameter dotplot?... a cluster? In a correlated 2 parameter dotplot, quantities of the cell properties, e.g. FSC (forward scatter) and the SSC (side scatter) intensity, are assigned to channels on the x- and y-axis. Each cell with a given FSC and SSC intensity is represented by one dot in the dotplot. Several cells with the same FSC/SSC combination share the same dot location. The number of cells in a channel can be represented by a color, according to a color scale. In a dotplot, subpopulations of cells with about equal properties appear as clusters. These clusters can be circled by drawing lines around them for more detailled analysis, e.g. in order to analyse the number of cells in this cluster. The process of analysing subpoulations is called "gating". 2 parameter dotplot of leukocytes with 256 x 256 channels. Side scatter intensity (SSC) is plotted versus the forward scatter intensity (FSC) in a linear scale. Colors (or grey levels) are a measure of the channel content. The dotplot shows the analysis of a blood cell sample. Preparation with Partec CyLyse. The circled cluster originates from the lymphocyte-subpopulation. FloMax Operating Manual Data Analysis 13

14 Introduction - Correlated 2 Parameter Dotplots What if data falls outside a histogram range? If parameter values fall outside of the histogram scale, e.g values are < 0.1 on a 4 decade logarithmic scale, FloMax will put them onto the histogram axis. These values will be taken into account e.g. in quadrant statistics. 2 parameter dotplot of a CD3/CD8 fluorescence analysis of lymphocytes with 256 x 256 channels. Note the logarithmic scale spanning a range of 4 decades. Data points below a relative value of 0.1 (circled) will reside on the axis. They are taken into account for quadrant statistics. FloMax Operating Manual Data Analysis 14

15 Introduction FloMax Start and Exit Start FloMax 1. Doubleclick on Partec FloMax Icon on the desktop or select Start-Programs - Partec FloMax from the Windows Start-button - FloMax starts and displays the Welcome dialog. FloMax Icon on the desktop. 2. Press OK - FloMax shows an initial page with histograms. Welcome dialog. Close Button Exit FloMax Make sure the actual document is saved. Click on application close button or select File-Exit - FloMax is finished. Initial histogram page. If changes to the actual open document have been made and not saved yet, FloMax asks to save the document before finishing. Refer to page 21 on how to save files. FloMax Operating Manual Data Analysis 15

16 Introduction Typical Steps for a Immunophenotype Analysis Open button 1. Open FCS file. View Page Layout 2. Select proper display. Right click on histograms 3. Select parameters to display. 4. Define lymphocyte gate. 5. Apply gate to histograms. 6. Compensate crosstalk. 7. Set quadrants. 8. Save FCS file. or Analysis Create Report 9. Print or Create Report. FloMax Operating Manual Data Analysis 16

17 Introduction Typical Steps for a Kinetic Analysis 1. Load FCS file. Open button View Page Layout 2. Select proper display. 3. Select a kinetic dotplot with the dynamic parameter (usually a fluorescence parameter) vs. the time parameter. 4. Enter square regions in equidistant positions to define a number of time slices. Right click on histogram 5. Select Analysis-Region Statistics - the region statistics displays the mean fluorescence values for each time slice. 6. Click Save to File in the Region Statistics dialog - the statistics results are saved to an ASCII text file. Kinetics dotplot with time slice regions. Analysis Region Statistics 7. Invoke e.g. Excel and open the text file. 8. Select Insert Graph to generate a graph from the mean values column. 9. Save/Print the Excel document. Region Statistics dialog. Start Programs - Excel Once the time slice regions are defined, click Save in the Gating dialog - time slice regions can then be reload to analyse other flow files. It might be of help to use the Zoom function before defining the time slice regions and possibly to select higher resolution data display for the dotplot for more precise entry. If data was acquired without a time parameter, the Analysis - Calculate Parameters feature of FloMax allows to generate an event number or pseudo time parameter that can be used for kinetic analysis, if the acquisition took place with a constant particle rate. Kinetics plot generated with Microsoft Excel. FloMax Operating Manual Data Analysis 17

18 Introduction Typical Steps for DNA Cell Cycle Analysis 1. Load FCS file. 2. Perform Cell Cycle Analysis. 3. Print Results. Typical cell cycle analysis analysis based on a numerical fit: G1, S, and G2/M phase of a CHO cell line. FloMax Operating Manual Data Analysis 18

19 Introduction Typical Steps for Microorganism Analysis 1. Load file. 2. Define microorganism regions. 3. Save file. 4. Print Results. Typical microorganism analysis example: Bacillus subtilis spores (red) and vegetative cells (blue). FloMax Operating Manual Data Analysis 19

20 Opening a Flow Cytometry Data File Open function The open function reloads the results which have been saved in a file before. Click on Open button in the toolbar - the Open dialog box appears. or Open button Click File in the menu bar - the file menu appears. Click Open... in the file menu - the Open dialog box appears. In the Look in field, select the folder where the file is located. Click on the file to be opened. The standard extension is.fcs. Click Open button - the file is opened and the results are displayed. Open dialog box or Click the Previous File or Next File button in the toolbar - the actual file will be closed and the previous/next file in the current folder in alphabetical order will be opened. Previous/Next File In case you forgot where the file was located, you may use the Windows search function in the Start menu. The file data format must be according to flow cytometry data standard FCS 2.0. You can also use the optional barcode reader to facilitate the selection of files, e.g. by using the sample barcode as filename. FloMax Operating Manual Data Analysis 20

21 Saving an Analyzed Flow Cytometric File Save function The save function saves the results in a file with the name given in the title bar: Click on Save button in the toolbar - the file is saved. or Click File in the menu bar - the file menu appears. Click Save in the file menu - the file is saved. Save button A previously opened file is overwritten by the Save function. Generally, it is recommended to use Save As function which always asks for a file name. In case no file name was specified yet, e.g. for a new document, the Save As dialog box appears (see below). Save As function The Save As function will ask for a file name before saving: Click File in the menu bar - the file menu appears. Click Save As... in the file menu - the Save As dialog box appears. In the Save in field, select the folder where the file is to be saved. In the File name field, enter a file name. Do not enter the extension - the file automatically will get the extension.fcs used for flow cytometry standard files. Click Save button - the file is saved. Save As dialog box Make use of different folders. E.g. use a different folder for each month or for each series of samples. This will help you when later locating your data. You may use long file names as supported by Windows (up to 256 characters). However, if you plan to use other software for further analysis (e.g. WinMDI), it might be possible only to recognize 8 character long filenames. You can directly save a file to another computer in your computer network by selecting the appropriate network folder. Refer to Windows manuals for details. Using Sample Barcodes You can also use the (optional) barcode reader to specify e.g. sample bacodes as file names. Barcodes can be generated by optional barcode software. FloMax Operating Manual Data Analysis 21

22 Printing the Histogram Page Print Now button. Printing the histogram page To print the actual histogram page: Make sure the file is saved. Click the Print Now button in the toolbar or Click File in the menu bar - the file menu appears. Click Print Now... in the file menu toolbar - the histogram page is printed or Click File in the menu bar - the file menu appears. Click Print... in the file menu - the Print dialog box appears. Check the printer settings and click OK - the histogram page is printed. Print dialog. The status indication in the taskbar notifies an ongoing printing process. Printing status indication. You can select a portrait (smaller) or landscape (larger) printout by selecting a paper orientation after clicking on the Properties button in the Print dialog box. Wait for the printing to complete before starting a new acquisition. Acquisition during printing may cause reduced counting accuracy. Print Preview In order to display a preview of the printed page on the computer display, a print preview can be selected. Click File in the menu bar - the file menu appears. Click Print Preview... in the file menu - the print preview appears. Move the Zoom glass by the mouse and click on regions of interest to see more details. Click on Close button to finish the preview. You may also use the report generator to print data according to your report form. In the reports, you can use annotations and your laboratory logo. Printing is performed with higher resolution than the display screen. Printing allows to achieve high resolution histograms and dotplots, e.g. with more than 1000 channels. FloMax Operating Manual Data Analysis 22

23 Printer Setup Change printer settings For permanently change of the printer settings, e.g. to select another paper orientation, click File in the menu bar - the file menu appears. Click Print Setup... in the file menu - the print dialog box appears. Make sure the selected the printer is the one which is actually connected. Printing is tested and guaranteed to work well for a series of printers (see appendix). Therefore, all these printers are recommended. However, printing works on almost any printer supporting Windows. Most printers require the installation of a printer driver software. The printer driver usually can be found either a) on disks provided by the printer manufacturer or b) on the Windows CD. Click the Properties button. Change the printer settings. Click OK. FloMax Operating Manual Data Analysis 23

24 Selecting the Page Layout Selecting the page layout The page layouts allow a quick change or setup of the histogram page. Here, the number of histograms on the page can be selected. The page layout also allows limitation of the analysis and display to a given range of particles, e.g. the first lymphocytes. Selecting a predefined page layouts Click View in the menu bar - the view menu appears. Click Page Layout... in the view menu - the Page Layout dialog box appears. To select the number of plots (histograms or dotplots) on the page, click on one of the page layout buttons. Selecting a range for data analysis and display To limit the data display and analysis to a range of events (cells), click on one of the Event Range options: All Events: all events of the file used. After selection of the basic layout, the type of the histograms (e.g. 1 parameter / 2 parameter) and parameters on the histogram axis can be selected by right-clicking onto the histograms. Alternatively, ready prepared histogram templates with all selections can be used as well. When switching to a lower number of histogram plots, the first plots of the page will be displayed. E.g. 8 initially plots: [1] [2] [3] [4] [5] [6] [7] [8] 3 plots: [2] [1] [3] 1 plot: [1] Events in Counting Volume: Only events in the counting volume are used (requires volumetric absolute counting from a Partec flow cytometer). First 10000/20000 Events: The first 10000/20000 events of the file are used. Preselected Events: Events between the two limits below (from-to) are used. Event Count Gate: A previously defined gate can be specified to define the event range for a gate. Cannot be used together with Events in Counting Volume. Select a gate to use a given number of subpopulations of cells. Example: Exactly lymphocytes can be analysed and displayed in cases the file contains data from more lymphocytes. Click OK - the new page layout is selected. FloMax Operating Manual Data Analysis 24

25 Selecting Histogram Properties For each of the histograms on the page, its type and properties can be individually changed. Open the histogram property box Right click on a histogram or doubleclick on a histogram - the property box for that histogram is opened. Select the histogram type To select a 1P histogram or a 2P dotplot, click on one of the radio buttons: 1P Histogram 2P Dotplot Change the histogram properties See the following pages about the properties and settings. You may create a histogram template once with all the histogram settings optimized for your application. To create a histogram template, select type and properties of the histograms and save it with File-Save As..., as a regular FCS file without data. To use a histogram template for a new acquisition, open this template (FCS file without data) before starting an acquisition. You may also load a usual FCS file with data and use the layout stored within the file as a template. To apply the settings, click Apply - new settings are applied to the histogram and the histogram property box stays open. Click Close - new settings are applied to the histogram and the histogram property box is closed. You can simply switch to the property box of another histogram by right clicking onto that histogram while the property box remains open. Black plot selection marks show which plot is actually selected. FloMax Operating Manual Data Analysis 25

26 Histogram Properties In the parameter selection list, only parameters which were (or are) stored in the previously loaded file or enabled for acquisition can be selected. Gate Parameter Click on the little arrow at the parameter-box - a list of the parameters appears. Click one of the parameters. Label Click on the little arrow at the label-box - a list of labels appears. Click one of the labels. or Enter a label for the parameter by keyboard. Color (1P histograms only) Click on the little arrow at the color-box - a list of available colours appears. Click one of the colors. Channels Click on the little arrow at axis channels-box - a list of available channel resolutions appears. Click on the desired number of channels. Parameter Name Parameter Label Location of parameter names and labels In case of a linear scale, the numbers on the histogram axis correspond to channel numbers. In case of logarithmic scaling, the axis always cover a range up to Automatic Rescale Click on this to enable/disable an automatic vertical rescaling during acquisition or rerun from file. y-axis Maximum (1P histograms only) Enter a number to define which y-scale will be used for the histogram. z-axis Maximum (2P histograms only) Enter a number to define which channel content will be used for the last color of the color scale. Optimize To optimize the channel content scale according to the histogram or dotplot, click this button. Optimize All All histograms are scaled optimally at once by one click to this button. FloMax Operating Manual Data Analysis 26

27 Histogram Properties z-level (2P dotplots only) Enter a number to define which is the minimum channel content to be displayed. You may use this to separate subpopulation clusters visually. Gate To activate a gate for the histogram or dotplot, click on the little arrow at the gating-box - a list of available regions and gates appears. Click on the desired region or gate. Color Gating Switches between color gating display and z- level-color display. Scatter dotplot: z-level = 0 Scatter dotplot: z-level = 6 FloMax Operating Manual Data Analysis 27

28 Analysing Channel Contents Analysing Channel Positions and Content Move mouse cursor across a one parameter histogram or two parameter dotplot. - the mouse cursor changes to a crosshair and the status line displays information on the channel number, position, and content. crosshair cursor Meaning of Channel Information a) One parameter histograms Example: x=166 (39.2) count = 198 channel number: 166 relative intensity value: 39.2 count: 198 b) Two parameter dotplots Example: x=171, y=64 (47.0, 1.0) count = 2 Channel Analysis in one parameter histogram. x-channel number: 171 y-channel number: 64 x-relative intensity value: 47.0 y-relative intensity value: 1.0 count: 2 Channel Number The range of channel numbers depends on the axis resolution set for the histogram. Typical values are or For a log scale channel values do not correspond directly to intensity values. Relative Intensity Value The range of relative intensity values is [0 1000] independent of the scale or histogram resolution used. These values takes into account the scale used (linear, 3 or 4 decade logarithmic). For log scale histograms, relative intensity values correspond to the axis values. Count The count is the channel content or number of particles of the channel currently addressed. Channel Analysis in two parameter dotplot. Channel contents are not displayed for dotplots which show color gated data. Make sure color gating is switched off for the plot before analysing channels. In case of a linear scale, the numbers on the histogram axis correspond to channel numbers. In case of logarithmic scaling, the axis always cover a range up to FloMax Operating Manual Data Analysis 28

29 Absolute Cell Counting / Analysing the Particle Concentration The Partec True Volumetric Absolute Counting system allows to analyse the concentration of cell subpopulations during or any time after acquisition, e.g. by defining or adjusting regions and gates, provided a count phase has been completed. FloMax displays the particle concentrations in the analysed tube. Remember to take into account possible dilution factors. For data for which a volumetric absolute count has been performed, concentration results of all or a subset of cells are displayed. The concentration, i.e. the number of all analysed particles per ml is shown in the result line on top of the histogram page. The result line also displays the total number of analysed particles. The concentration of particle subpopulations is displayed in the region statistics field for each region or gate. For quadrants, the concentration of particlesubpopulations is additionally directly displayed in the quadrants. CD34+ volumetric absolute counting without reference beads. Two methods can be used for absolute cell counting i) a reference method, ii) Partec s unique volumetric absolute counting method. Direct display of the concentration of cell subsets is supported for method (ii). For (i), the ratio of counts from the subpopulation and the reference beads has to be multiplied with the specified bead concentration. FloMax Operating Manual Data Analysis 29

30 Rerun From File Function Rerun button. For huge files with several hundred thousand events and complex displays, the rerun function can take some tens of seconds, depending on the computer speed. FloMax displays a progression percentage in the status line. For a simpe replay animation, data from the active file can be rerun in the sequence of acquisition (but faster). This can be a simple means to check the stability of the acquisition, e.g. if peaks shifted over time. It might also help to demonstrate how an acquisition proceeds. Click Rerun button in the toolbar or Select Analysis Rerun From File in the menu or Simultaneously press Ctrl and R on the keyboard - data is rerun and reclassified into the histograms as during acquisition. FloMax Operating Manual Data Analysis 30

31 Zoom Function Zoom button. Using the Zoom Function For a closer look to the data plots, a zoom function is provided. The Zoom function magnifies an area of interest. It e.g. helps to precisely define regions when using histogram pages with 6 or 8 plots. 1. Click the Zoom-button in the toolbar or Press F2 on the keyboard or Select View-Quick Zoom in the menu - the screen zooms to a portion of the histogram page. 2. Click on the scroll bars to move to the plot of interest. Zoomed histogram page for precise region entry. The Zoom function can also be used to generate high quality copy & paste results, e.g. for publications. 3. Perform the analysis which requires a closer look. 4. Click the Zoom-button in the toolbar again - the screen reverts to normal size. For more zooming, select View-Zoom+ and Zoom- in the menu. FloMax Operating Manual Data Analysis 31

32 Gating - Defining Regions Why Gating? By gating, subpopulations of particles can be analyzed. The display and statistical evaluation can be reduced to a subpopulation defined by a gate. Typically, a gate is a simple range in a 1P histogram or a polygon region in a 2P dotplot. Gates can also consist of a logical combination of ranges, polygon regions, and quadrants. To perform gating, two steps are required: 1. Defining gates (or regions). 2. Applying gates to histograms. Defining regions Click on one of the gating region buttons in the toolbar or click on the G-button - the gating dialog appears. Entry of a polygon to define the lymphocyte cluster. Polygon entry in a 2P dotplot 1. Click the polygon button. 2. If desired, specify a name for the polygon in the Region entry of the gating dialog. If not specified, default names as R1, R2 etc. are used. 3. Click the first point of a polygon in a 2P dotplot. 4. Draw the polygon, a) click once for every corner of the polygon or b) draw a region while keeping the mouse key pressed. 5. To close the polygon, a) click in the box around the first point or b) press the right mouse button - the polygon is closed and defined. Quadrant to define lymphocyte subpopulations. Quadrant entry in a 2P dotplot 1. Click the quadrant button. 2. If desired, specify a basic name for the quadrants in the Region entry of the gating dialog. If not specified, standard names as Q1-Q4, QA1- QA4 etc. are used. 3. Move quadrant cross to desired point in a 2P dotplot and enter by mouse click - four quadrant regions are defined. Range in one-parameter histogram. See next page on how to apply gates. Range entry in a 1P histogram 1. Click range button. 2. If desired, specify a name for the ranges in the Region entry of the gating dialog. If not specified, standard names as RN1, RN2 etc. are used. 3. Move cursor to left border of desired channel range in a 1P histogram. Enter left border by mouse click. 4. Move to right border of range. Enter right border by mouse click. - the 1P range is defined. FloMax Operating Manual Data Analysis 32

33 Gating - Applying Gates Applying gates or regions to histograms 1. Right click the histogram for which a gate is to be applyed - the histogram property box is displayed. 2. Click the little arrow in the gate drop down list and select the gate or region. 3. Press OK - the gate is applied and the histogram property box is closed. If a gate is applied to a histogram, this is displayed at the top of the histogram. Removing gates or regions from a histogram 1. Right click the histogram for which a gate is to be applyed - the histogram property box is displayed. 2. Click the little arrow in the gate drop down list and select the <None>. 3. Press OK - the gate is removed from the histogram and the histogram property box is closed. 2P dotplot of FL1 - FL2, gated by a polygon region "Lympho" defined in the FSC - SSC dotplot. Gating can be applied during or after an acquisition: online and offline gating. In any case, all events are saved. This allows to completely reanalyse the measurement, starting from the original data. FloMax Operating Manual Data Analysis 33

34 Gating - Logical Gates Defining logical gates If not already opened, open the gating dialog by pressing the gating button in the toolbar. Click Logical Gates - the Logical Gates dialog appears. In the Logical Gates dialog, select a name for the gate and assign to it a logical gate expression. As an example to define a gate for particles belonging to R1 and R2 as well, enter G1=R1 AND R2 By using the operands AND, OR and NOT combined with the brackets ( ), any logical combination of regions is possible. You may select regions from the regions list and use the buttons to enter operands and brackets. Logical Gates dialog Press the [ = ] button to validate the gate definition. Note: CLR clears the entry for a gate. It is not neccessary to define gates for simple regions, e.g. G1=R1. Simple regions can be directly applied as a gate. Press OK the logical gates appear in the left gates list and can now be used for gating. FloMax Operating Manual Data Analysis 34

35 Gating - Saving and Loading Gates and Regions Saving gates and regions A complete set of gates and regions may be saved in a file. These gates can later be reload and applied to other measurements. 1. Press G= to open the gating dialog. 2. Click on Save... in the gating dialog - the Save As dialog box appears. 3. In the Save in field, select the folder where the gating file is to be saved. 4. In the File name field, enter a file name. It is not neccessary to enter the extension - the file automatically will get the extension.gat used for gating files. 5. Click Save button - the gating file is saved. Regions, ranges and quadrants are stored in a mathematical space spanned by the parameters. E.g. when quadrants are defined for a FL1-FL2 dotplot, they will allways appear in a FL1-FL2 plot, no matter where this plot is displayed on the screen. The quadrants will even stay defined if no FL1-FL2 plot is on the screen at all. Loading gates and regions 1. Press G= to open the gating dialog. 2. Click on Load... in the gating dialog - the Open dialog box appears. 3. In the Look in field, select the folder where the gating file has been saved. 4. Click on the gating file to be opened. The extension is.gat. 5. Click Open button - the gating file is opened - gates and regions are shown in the histograms. Saving a gating strategy protocol A complete gating strategy, consisting of gates, regions and the histograms where they are applied, can be stored within a histogram template, which is nothing else than a FCS file without data. To use this, 1. Prepare a histogram template with suited histogram types and parameters on the axis. Tip: Reload a measurement. 2. Define regions and gates and apply these to the histograms. 3. Save this file by the File - Save As... function as a FCS file. Loading a gating strategy protocol Load a histogram template or measurement which contains a gating strategy. This can be used for new acquisitions. FloMax Operating Manual Data Analysis 35

36 Gating - Moving, Resizing, and Deleting Regions Moving regions 1. Select the polygon or quadrant region to move: click with the mouse near the region or quadrant center - a selection frame appears. 2. While the mouse is above the selection frame, a double cross is displayed. While holding down the left mouse button, move the region. 3. Release the mouse button - the region is moved and region statistics are updated. tracker Resizing polygon regions 1. To select the polygon or quadrant region to resize, click with the mouse near the region or quadrant center - a selection frame appears. 2. While the mouse is above one of the square trackers, a double arrow is displayed. While holding down the left mouse button, move the tracker. 3. Release the mouse button - the region is resized and region statistics are updated. Deleting regions 1. To select the polygon or quadrant region to delete, click with the mouse near the region or quadrant center - a selection frame appears. 2. Press the Del key on the keyboard - the region or quadrants are deleted. or 1. Open the gating dialog with the gating button. 2. In the Gates drop down list, select the region, quadrant, or range to delete. 3. Press the Delet button in the gating dialog - the region is deleted. Deleting logical gates To delete all gates and regions, press Clear All in the gating dialog. 1. Open the gating dialog with the gating button. 2. In the Gates drop down list, select logical gate to delete. 3. Press the Delet button in the gating dialog - the logical gate is deleted. FloMax Operating Manual Data Analysis 36

37 Gating - Obtaining Statistics For each of the polygon regions, quadrants or 1P ranges defined, various statistical information can be displayed. Select display of region statistics Make sure View - Region Statistics is checked in the menu - results are displayed on the lower part of the histogram page. Use the scroll bar on the right side or reduce the size of the histogram window to display results at the lower part of the histogram page. Select statistics information To specify which results are to be displayed 1. Select Analysis - Region Statistics... in the menu - the Region Statistics box appears, containing the regions defined so far. 2. Select the results to show/to hide. 3. Click Close - the Region Statistics box is closed and the results are displayed on the lower part of the histogram page. For log-scale data, geometrical mean, mean and CV-calculations are performed after transforming all data to the linear scale first. See mathematical appendix for details of calculations. Region Statistics Results Region name of the region Gate name of the gate in effect. <None> for ungated results. Ungated total number of cells or particles belonging to the region. Count number of particles belonging to the region, gated by the given gate. Count/ml number of particles in the region per ml. Only available if absolute counting has been performed. Gated, if a gate is specified. % Gated percentage of particles related to the number of particles in the given gate, if a gate is active. % Total percentage of particles related to all particles. GMean geometrical mean of the region in x- or y- direction. Mean mean of the region in x- or y-direction ("center of gravity"). CV% relative coefficient of variation (standard deviation divided by mean value) of the region in x- or y- direction. FloMax Operating Manual Data Analysis 37

38 Gating Saving Region Statistics to a File Saving Region Statistics to a File The region statistics results can be saved to a text file (ASCII) for export and simple further evaluation by other software, e.g. Word or Excel. 1. Select Analysis - Region Statistics... in the menu - the Region Statistics box appears, containing the statistical results. 2. Select the results of interest. 3. Press Save To File... - a Save As dialog appears. 4. Specify filename and folder for the export file and press OK - Region Statistics is saved in an tab-text ASCII file. The Save to File function generates an ASCII text file in tab-text format. This format can be imported as a data table by almost any other software. FloMax adds the extension.txt to the exported file. To import the region statistics from the ASCII text file, select the appropriate import function in the other software. Be sure to specify the correct extension (.txt). Select a tab-text format. FloMax Operating Manual Data Analysis 38

39 Gating Exporting a Gated File Exporting a Gated File For other evaluation software it might be necessary to work on data which is already analysed to some degree, e.g. gated or compensated. E.g. an external cell cycle analysis software could be used to evaluate subpopulations contained in the FCS data set, but might itself not offer gating options. For this reason, FloMax offers the export of gated or compensated files. 1. Define the gates, as decribed further above. 2. Select FIle-Save Options - the Save Options dialog appears. 3. In the gate drop list, select the gate to use for the export. 4. Enable the Gated by checkmark. 5. Press OK the Save Options dialog is closed. 6. Select File-Save As - the Save As dialog appears. 7. Select a folder and a filename and press OK - Together with a normal FCS file with all data, a second file with the same name plus an extension.a is saved. The.A -file is a regular FCS file but only contains the gated data. To import gated data by other software, select FCS as data type and.a (for Analysis) as file extension. Example Filename selected in Save As is EAT two FCS files are generated: 1. EAT.FCS, containing all data. 2. EAT.FCS.A, only containing gated data. The generation of a second gated or compensated file as specified in Save Options will be active unless explicitely deactivated. This facilitates rountine operation when e.g. always having to save gated files. When no more required, be sure to deactivate the generation of a gated/compensated file by clearing the save options. FloMax Operating Manual Data Analysis 39

40 Gating - Color Gating Color Gating Color gating is used for simple visual association of cell-clusters in different dotplots. Generally, by color gating, a unique color is associated to regions or gates, called color gates. Cells belonging to a color gate, are then displayed with the associated color. However, things may become more complicated when a cell belongs to more than one color gate. FloMax offers two different color gating modes for these cases: Priority Color Gating MultiColor Gating Priority Color Gating Events in color gates are displayed on top of each other, according to the priority: color gate 1 > color gate 2 > color gate 3. The color for each of the three color gates can be freely selected. Any range, region, quadrant, or logical gate (logical combination of ranges, regions, quadrants) can be associated with color gate 1, 2, or 3. MultiColor Gating All combinations of color gates are displayed by an independant color. All colors are freely selectable. MultiColor Gating. Circle overlaps are associated with an independant color each. The color gating setting is stored within FCS files or, together with the regions and gates, in separate gating files, which can be applied easily. Gating Procedure 1. Define polygon-regions, quadrants, and oneparameter ranges in dotplots and histograms, e.g. R1 = Lymphs, R2, Q1...Q4, QA1...QA4, RN1. 2. Define logical gates, as required, e.g. G1 = Lymphs AND Q2. 3. From the Gating dialog, click Multi Color Gating... - the Color Gates dialog appears. 4. Select the color gating mode: MultiColor. 5. Select the regions/gates to be used as color gates. 6. Change color of the color gates, if required. 7. Click OK - the Color Gates dialog is closed. 8. Make sure Color Gating is enabled for each dotplot you want to use color gating. - the events are displayed with the corresponding gate color. FloMax Operating Manual Data Analysis 40

41 Gating - Color Gating Selecting/Deselecting Color Gating for all plots Select View-Color Gating all plots on the page are displayed according to the new setting. Selecting/deselecting Color Gating for all plots. Selecting Color Gating for individual histograms 1. Right click onto dotplot for which to change color gating display the histogram property box is opened. 2. Select/deselect the Color Gating flag. 3. Press Close the dotplot is displayed according to the selection. Selecting/deselecting Color Gating for individual plots. Color Gating on a four color analysis. Lymphocytes appear red, monocytes green, granulocytes blue. FloMax Operating Manual Data Analysis 41

42 Crosstalk Compensation - Overview You can compensate any parameters, e.g. FL2 to FL1, FL3 to FL1, FL1 to FL3, and SSC to FL4. You can correct and repeat the compensation anytime, even after reloading a measurement file from disk. Why Crosstalk Compensation? Crosstalk compensation is a means to compensate for spectral overlap of fluorescent dyes, when two or more dyes are used. FloMax provides a software based mathematical compensation. This allows to take into account the crosstalk between any (of N) parameters, i.e. not only the neighbored colors. The algorithm used is correspondingly called N-color compensation. Unique advantages of a software based algorithm are: a) Compensation settings can be corrected any time during or after the acquisition without need to rerun samples. b) Compensation is reproducible and does not depend on possibly variing amplifier noise and offsets in the millivolt region which are usually difficult to control for the first decade on a 4 decade log scale. FloMax offers graphical entry of crosstalk values, thereby facilitating the compensation setup. Graphical entry of crosstalk entries. FloMax N-Color software compensation requires the FCS files to be stored without previous compensation. If you are not using a Partec flow cytometer, switch off your instrument compensation in order to use FloMax compensation. FloMax provides stepwise compensation, e.g. starting with FL1 FL2, then FL2 FL3, and so on in the order of strength of crosstalk. Compensating one parameter against another can influence the distributions of all parameters (all histogram plots) included in the compensation matrix. FloMax Operating Manual Data Analysis 42

43 Crosstalk Compensation - Setup or The traceline corresponds to the values, which will be compensated to zero (or onto the axis). All events below the traceline will also be compensated to zero. Different compensation modes can be specified. Log is the preferred mode taking into account if and what log scales are used. Lin always deals with data as if it was acquired with linear amplification. This mode does not change the general cluster shape. However, it this mode is not strictly correct. Only use in special cases. Random Bias can be selected to redistribute data otherwhiles falling onto the histogram axis graphically (see notes for details). When starting a FloMax controlled instrument, the previously set compensation values are active. Warning Be sure not to overcompensate too much. The result could be an underestimation of dimly stained positive cells. Set up the Compensation 1. Select a display with dotplots showing the parameters to compensate against each other, e.g. FL1-FL2, FL2-FL3. 2. Open the compensation box: a) Click the compensation button on the toolbar or b) Select Analysis - Compensate Crosstalk... - the Compensation toolbox appears. 3. To work with new settings, press Clear All... in the compensation toolbox. 4. Click in the 2P dotplot showing the largest crosstalk, e.g. FL1-FL2. If data was already compensated by FloMax, press Undo to show the compensation tracelines. 5. Enter slope (compensation percentage) and intercept (autofluorescence values): a) Method 1: Graphically by compensation trace lines: Move the mouse cursor above the square line trackers - the cursor is changed to a double arrow. With the left mouse button pressed, move the the trace line to set the slope and intercept. The corresponding values are displayed in the compensation box. b) Method 2: Numerical entry of values. In the boxes Crosstalk of and to select the parameters from and to which crosstalk to compensate. Enter slope and intercept for this compensation. 6. Click Compensate Crosstalk. 7. If desired, select Random Bias. 8. Click the + and buttons left of the slope entry box to increase/decrease the compensation value of the actual parameters the effect is directly visualized in the plots. 9. Repeat steps 4-8 with the dotplots remaining for compensation. 10. To close the compensation toolbox, click Close. Revert to Uncompensated Data for the Actual Dotplot Click Undo - the uncompensated of the actual dotplot data is shown. To correct the compensation, repeat the steps above. Revert to Uncompensated Data for All Parameters Click Clear All the original uncompensated data is shown. To correct the compensation, repeat the steps above. To keep the data uncompensated, close the compensation toolbox by Close. FloMax Operating Manual Data Analysis 43

44 Crosstalk Compensation Saving and Loading Saving and Loading Compensation Settings Click Save... / Load... in the compensation dialog and specify a filename and folder in the Save As / Open dialog box: - Compensation settings will be saved / loaded from disk. FloMax Operating Manual Data Analysis 44

45 Crosstalk Compensation Exporting a Compensated File Why Exporting Compensated Files? For other evaluation software it might be necessary to work on data which is already compensated. E.g. an external software could be used for special 3D data display, but might not offer a compensation function. For this reason, FloMax offers the possibility of exporting compensated files. Exporting Compensated Files 1. Compensate the data, as decribed above. 2. Select File - Save Options - the Save Options dialog appears. 3. Enable the Compensated checkmark. 5. Click OK the Save Options dialog is closed. 6. Select File - Save As - the Save As dialog appears. 7. Select a folder and a filename and press OK To import compensated data by other software, select FCS as data type and.a (for Analysis) as file extension. Files written this way contain compensated data in FCS format. These files can be read by FloMax, but compensation cannot be corrected or undone. If you do not need to export data, it is recommended to save files using FloMax standard FCS format. Opening standard format files with FloMax will show compensated data but allow to revert to the original data, since FloMax saves original data plus the compensation matrix. - Together with a normal FCS file with all data, a second file with the same name plus an extension.a is saved. The.A -file is a regular FCS file but contains compensated data. Example Filename selected in Save As is Apoptosis two FCS files are generated: 1. Apoptosis.FCS, containing original data plus compensation settings. 2. Apoptosis.FCS.A, containing compensated data. The generation of a second gated or compensated file as specified in Save Options will be active unless explicitely deactivated. This facilitates routine operation when e.g. always having to save compensated files. When no more required, be sure to deactivate the generation of a gated/compensated file by clearing the save options. FloMax Operating Manual Data Analysis 45

46 Crosstalk Compensation - Notes FL1 -> FL2 = 10% with autofluorescence mit Autofluoreszenz FL1 -> FL2 = 20% ohne Autofluoreszenz without autofluorescence Original data example before compensation with appropriate compensation line (schematical). Why is the Cluster Size Increased by Compensation? Correct compensation tends to increase the cluster size graphically on a logarithmic scale (see example on the left). Even if correct in a strict mathematical sense, the increase of cluster size due to compensation is often undesired especially when operators are familiar with less precise electronic compensation techniques. Example Assume a fluorescence plot on a 4 decade logarithmic scale as on the left. Your uncompensated FL1 negative and FL2 positive cluster extends from fluorescence values 3 to 10 in FL1 and 30 to 100 in FL2. Its mean value is at (FL1, FL2) = (7.3, 70). A correct compensation would be 10% (slope). Accordingly, the FL1 mean value is compensated to: FL1-10% x FL2 = % x 70 = 0.3. The right border of the cluster is compensated to 10-10% x 70 = 3. The left border of the cluster is compensated to 3-10% x 70 = - 4. How would this compensated cluster appear on a 4 decade log scale? The cluster mean value will be presented about in the middle of the first decade (value 0.3). Compensated data showing increase of cluster size. The effect of cluster size increasing has nothing to do with the method of compensation, e.g. by analogue means or by software. The right border (value 3) will appear about in the middle of the second decade. The left border (value 4) cannot be displayed on the log scale since it is non-positive. However, negative values will be limited to the minimum displayed value on the axis, which is 0.1 for a 4 decade scale. The cluster will consequently range from the axis to the middle of the second decade. It's size is increased significantly. Generally, this effect will grow with increasing compensation values. FloMax Operating Manual Data Analysis 46

47 Crosstalk Compensation - Notes What is Random Bias? Random Bias is a means to simulate electronic noise and offsets present in traditional analog compensation circuits. Original uncompensated data with suggested setting of compensation traces. Q1 Q3 Q2 Q4 Slightly overcompensated data without random bias. Q1 Q3 Data from above with with random bias. Data is redistributed in the first decade of the log scale. Quadrant results will be similar with or without use of random bias. Warning Especially when using random bias, be sure not to overcompensate too much. The result could be an underestimation of dimly stained positive cells, which will appear in the negative distribution. Q2 Q4 How to use Random Bias? In order to graphically present data as more narrow clusters, data may slightly be overcompensated by setting the trace lines slightly above the "negative-positive" clusters. By compensation, mathematically, this cluster will be completely moved to the axis. By adding a Random Bias to the data, a small artificial offset with Gaussian-like distribution, this cluster will be moved back from the axis to a narrow cluster near the axis. The effect of all this on region and quadrant statistics usually is small. However, if you prefer a mathematically correct compensation, it is recommended to turn off the Random Bias. Should I Use Random Bias? If you want to work with a mathematically correct compensation, no. If you need to display data of negative clusters with a small distribution in the first decade, yes. - The influence onto the quadrant statistics usually is small. How is Data Analysed on the Axis? Data on the axis will be included in quadrant analysis. What is the Influence of Gain Values onto Compensation Values? Required compensation slopes heavily depend on the gain values used for the optical detectors. E.g. when using doubled amplification for FL1 (usually by increasing gain by 50) in the left example, compensation would have been set to 20% instead of 10%. Can I Analyse without Compensation? If or if not spectral crosstalk compensation is necessary, depends on the characteristics of the used fluorochromes. Generally, for a two color analysis, the cluster percentages and concentrations can always be analysed without compensation as well. In cases where quadrants do not separate the clusters, polygon regions can be used. For multiparametric data, analysis without compensation can be difficult if not impossible. FloMax Operating Manual Data Analysis 47

48 Cell Cycle Analysis or Cell Cycle Analysis should be performed on a single histogram layout. Linear amplification must be used for the DNA parameter. Press Clear All to revert to the original display without analysis. drawing curve: numerical fit (sum of G1, S and G2/M fits). dottet points: experimental blue: G1 phase fit red: S phase fit green: G2/M phase fit See mathematical appendix for details about cell cycle algorithms. Performing a Cell Cycle Analysis 1. Make sure the DNA distribution showing the cell cycle is shown on a single one parameter histogram page. A linear scale is required for Cell Cycle Analysis. 2. Click the Cell Cycle Analysis button in the toolbar or Select Analysis - Cell Cycle Analysis - the Cell Cycle Analysis box is opened. 3. Select a fit method. a) For a complete numerical fit of a single cell cycle, click Fit One Cycle. or b) To perform only a Gaussian fit to G1 and G2/M data and analyse S-phase data "as-is" (without fit), click Fit G1/G2M only. - a numerical fit to the experimental data is performed, according to a mathematical model (see mathematical appendix for details). The numerical result is shown in the Cell Cycle Analysis Box. A graphical representation of G1, S, and G2/M phase is displayed together with the experimental data. Channel contents are now displayed as dots. 4. Click Close to close the Cell Cycle Analysis box - Results stay displayed in the histogram. Quantitative Cell Cycle Results G1% percentage of G1 phase cells related to all cells in the cell cycle S% percentage of S phase cells related to all all cells in the cell cycle G2M% percentage of G2/M phase cells related to all cells in the cell cycle CV%G1 relative coefficient of variation of G1 peak (relative width of Gaussian fit peak). CV%G2M relative coefficient of variation of G2M peak (relative width of Gaussian fit peak). MnG2M mean channel value of G2/M peak (mean of Gaussian fit peak). MnG1 mean channel value of G1 peak (mean of Gaussian fit peak). G2M/G1 ratio of MnG2M and MnG1. The ratio is 2 for normal cell cycles. ChiSqu. measure of the variation between experimental data and the fitted mathematical model. The smaller ChiSqu., the better the fit. FloMax Operating Manual Data Analysis 48

49 Peak Analysis Numerical Fit Method or Numerical Peak Analysis should be performed on a single histogram layout. For the calculations, a linear amplification is assumed. Press Clear All to revert to the original display without analysis. drawing curve: numerical fit (sum of peak fits). solid area: peak fits dottet points: experimental measurement data See mathematical appendix for details about peak analysis algorithms. By the numerical fit method a mathematical model (sum of bell-shaped Gaussian distributions) is fitted to the histogram data. In the model, each peak is represented by a Gaussian distribution with given position, width, and height. Performing a Peak Analysis 1. Make sure the data is displayed on a single one parameter histogram page. A linear scale is assumed for numerical Peak Analysis. 2. Click the histogram selection marks are displayed in the corners. 3. Click the Peak Analysis button in the toolbar or Select Analysis - Peak Analysis... - the Peak Analysis box is opened. 4. Click Fit Gauss Peaks: - a numerical fit to the experimental data is performed, according to a mathematical model (see mathematical appendix for details). The numerical result is shown in the Peak Analysis Box. A graphical representation of the fit is displayed together with the experimental data. Channel contents are now displayed as dots. 5. Press Close to close the Peak Analysis box - Results stay displayed in the histogram. Quantitative Peak Results Peak peak number (from left to right) Index relative peak position related to the first peak (or a reference channel) Mean mean channel position of the peak Area peak area (of Gaussian peak). Corresponds to the number of particles belonging to that peak. Area% percentage of peak area related to all sum of all peak areas CV% relative coefficient of variation of peak (relative width of Gaussian fit peak). ChiSqu. measure of the variation between experimental data and the fitted mathematical model. The smaller ChiSqu., the better the fit. FloMax Operating Manual Data Analysis 49

50 Peak Analysis Find Peaks Range Method & Batch Analysis By the find peaks range method, the position of the peaks in the histograms is analysed. For each peak, an appropriate left and right border is found. The peaks within these ranges are analysed. This method is also offered as a batch mode function, suited for the automated peak analysis of a series of files. Press Clear All to clear the Find Peak analysis. Performing a Range Method Peak Analysis 1. Click the histogram of interest selection marks are displayed in the corners. 2. Click the Peak Analysis button in the toolbar or Select Analysis - Peak Analysis... - the Peak Analysis box is opened. 3. Click Find Peaks: - left and right range markers are set for each peak. The range statistics (see region statistics) contains all relevant peak information as area, concentration, information about the position, and CVs. 4. Press Close to close the Peak Analysis box - Results stay displayed below the histogram in the region statistics field. Ranges automatically set by Find Peaks. Batch Mode Peak Find Analysis 1. Press the Peak Analysis button in the toolbar or Select Analysis - Peak Analysis Click Batch Analysis - the Select Output FIle dialog appears. 3. Select folder and filename for a text file to receive the results of all batched files. 4. Press Save. 5. Select the folder and a series of files which are to be analysed. Selection of output file for batch analysis. For data files with several histograms, batch mode analyses the first histogram in a file. Make sure this is the histogram of interest. Tip: To select a consecutive series of files, click on the first file in the list, then move to the last file for the series and click the last file while holding down the Shift key. To select/deselect an individual file from the current file list, click on the file while holding down the Ctrl-key. 6. Press Open all selected files are analysed one after the other. The result is written in form of a table to an tab-text ASCII file which can be read by almost any other software, e.g. Word or Excel. FloMax Operating Manual Data Analysis 50

51 Peak Analysis Histogram Export Why Exporting Histogram Data? For external cell cycle or peak analysis software that do not support loading of FCS files, it can be required to export a histogram in a simple channel-by-channel format. FloMax offers the Export Histogram function for this purpose. The number of channels (channel resolution) used for the export corresponds to the actual histogram resolution set in the histogram properties box. Exporting Histogram Data 1. Click the histogram to export selection marks are displayed in the corners. 2. Click the Peak Analysis button in the toolbar or Select Analysis - Peak Analysis... - the Peak Analysis box is opened. 3. Click Export Histogram - a Save As dialog appears. 4. Enter folder and filename for storage of the histogram. 5. Click Save - an ASCII tab text-file is generated with a list of the channel contents. This file can be read with many other software like graphics programs, Excel, or a text editor. Format of an exported histogram with channel numbers (left column) and channel contents (right column) as seen with Windows Notepad. FloMax Operating Manual Data Analysis 51

52 Ratio Analysis Purpose of Ratio Analysis Ratio Analysis allows to use the ratio of two measured values as a parameter. In FloMax, ratio analysis can be performed during acquisition or at analysis time, talking advantage of the built-in parameter calculation feature. Ratio analysis is used for ph-measurements or for spectral two-point analysis. Ratio Dialog. Performing Ratio Analysis 1. Select Analysis Calculate Parameters Ratio - the Ratio dialog appears. 2. If desired, in the Ratio Parameter box, select a name for the new ratio parameter (or use the default names EQ1, EQ2 etc.). 2. Select the parameters for ratio analysis. 3. Select the desired axis scale, e.g. log4 for fourdecade logarithmic. 4. Select the desired calculation base (math). Select on Linearized Values to take into account any log scales (reccomended). 5. Click [ = ] the ratio parameter is defined. 6. Click OK the Ratio dialog is closed. 7. Click in one of the histograms and select the ratio parameter for (one of ) its axis the ratio parameter is displayed and can be used as any other parameter. Ratio Analysis as a one dimensional projection (right) of a two parameter data set (left). The left peak of the histogram corresponds to the grean stained particles. The Ratio function scales the data such that a ratio of 1:1 is displayed around a value of 10, i.e. in the middle of a four decade histogram. Ratio calculation supports 16 bit (65536 channels) data resolution from Partec FCMs and employs double precision floating point arithmetics. This minimizes artifacs from rounding errors. For data acquired on a non-partec FCM with lower data resolution, e.g. 8 bit (256 channels) or 10 bit (1024 channels), rounding errors can possibly become visible. FloMax Operating Manual Data Analysis 52

53 Calculated Parameters Purpose of Calculated Parameters Calculations with parameters can be a very flexible tool for data analysis. Applications include linearizing log-scaled measurements. up- and down scaling of data when gain was not set to optimum. checking measurement kinetics by introducing a pseudo-time parameter. performing non-linear parameter transformations for better graphical visualization. calculating parameter projections to facilitate data analysis, e.g. for spectral analysis. Parameter Equations editor. Parameter calculations support 16 bit (65536 channels) data resolution from Partec FCMs and employs double precision floating point arithmetics. This minimizes artifacs from rounding errors. For data acquired on a non-partec FCM with lower data resolution, e.g. 8 bit (256 channels) or 10 bit (1024 channels), rounding errors can possibly become visible, depending on the formulas used. Defining Calculated Parameters 1. Select Analysis Calculate Parameters Formula - the equation editor appears. 2. Enter a name for the new parameter (or use the default name EQ1, EQ2 etc.) 3. Define the equation in the definition box right of the equal sign (=). Select functions from the drop function down list, arguments from the list of parameters, or the calculator keypad. 4. Select an axis scale for the new parameter (lin, log3, log4 for linear, 3 or 4 decade log scale). 5. Select the desired calculation base (math). Select on Linearized Values to take into account any log scales in the original data (recomended). 6. Click [ = ] the calculated parameter is defined. 7. Click OK the Equation editor is closed. 7. Click in one of the histograms and select the calculated parameter for (one of) the histogram s axis the calculated parameter is displayed and can be used as any other parameter. Math on Channel Values means the mathematical operations appear directly on the channel number associated with the actual parameter. Math on Linear Values means the operations appear on values that represent the original physical quantity, e.g. the light intensity. The latter takes into account possible logarithmic amplifications by linearizing (delog) the data before performing the calculation. This is the recommended method. FloMax Operating Manual Data Analysis 53

54 Calculated Parameters Examples Pseudo Time parameter unrevealing slight unstability during the run of a control sample. Event Number or Pseudo Time Parameter Listmode data contains information about the sequence particles were analysed. Therfore an event number parameter can be generated, which contains nothing more than the unique sequential number of a particle assigned when it was analysed. When assuming a constant particle rate, this particle number or event number can be used to define a time sequence parameter. Since this parameter does not strictly define a time due to variations in the particle rate (even if only statistically) and in order to differentiate it from the real time parameter that can be selected during acquisition, we call this pseudo time parameter. Nevertheless it can be of help for kinetic analysis or to check the stability throughout the course of a measurement. To use the pseudo-time parameter, select PTIME = EVNO() * 2.0 as a parameter equation. The multiplication of a factor can be used to strech or shrink the pseudo time scale. Scaling of Parameters Frequently for purpuse of comparisons it is required to scale parameters such as to make a peak appear at a specific position. Usually this is done during acquisition by changing the photomultiplier high voltage or amplification gain. Once stored, the gain can no more be changed. However, data can be scaled. This powerfull tool can avoid rerun of samples in many cases. Original histogram with normal G1 peak at channel 232 (as analysed with the channel info function). Example A DNA histogram is to be scaled such that the first peak of the histogram appears at channel Locate the original position of the peak by moving the cursor on top of it and note down the channel value in the status bar on the bottom of the page (e.g. 232). 2. Define a new parameter with formula EQ1 = FL1 * 200 / Select this new parameter (EQ1) for the histogram - the histogram is now scaled with the peak appearing at channel 200. Histogram scaled to display normal G1 peak exactly at channel 200. FloMax Operating Manual Data Analysis 54

55 Calculated Parameters Examples For data acquired on a non-partec FCM with lower data resolution, e.g. 8 bit (256 channels) or 10 bit (1024 channels) only, rounding errors can possibly become visible for lin-log transformations. Lin Log Transformation Data acquired on a linear scale can be transformed to a log scale and vice versa. The most simple method is to use one of the predefined lin/log scales but scales with any (as well broken number) decade log scales are also available. Standard conversion lin lg3 lg4 1. Define a parameter as EQ1 = FL1 2. Select the desired predefined scale (lin, lg3, lg4). 3. Click [ = ] a new parameter with FL1 data on the transformed scale is generated. 4. Select the new parameter for the display the transformed parameter is shown and the axis description is adjusted to reflect the linear, 3- or 4- decade scale. Any conversion lin log 1. Define a parameter by using the transform to log function as EQ1 = TLOG(FSC,2.5) 2. Make sure a linear scale is selected. Original (left) and transformed FSC-SSC dotplot with 2.5 decade transformation for SSC. This kind of transformation can improve the visibility and definition of the lymphocyte cluster. 3. Click [ = ] a new parameter with FSC data in a 2.5 decade scaling is generated. 4. Select the new parameter for the display the transformed parameter is shown. For any but the predefined scales, the axis description is linear and shows channel values and no relative intensity values. FloMax Operating Manual Data Analysis 55

56 Calculated Parameters Function Overview Arithmetical expressions: * / - + Brackets: ( ) RATIO(p1,p2) Calculates the ratio of p1 and p2 and scales the data conveniently so a ratio of 1:1 appears at a relative value of 10. Formula: RATIO(p1,p2) = p1/p2 * R/2. LG10(p) Calculates the log to the basis 10 of p. POW(p,f) Raises p to the power of f. TLIN(p,f) Transforms p defined on a log scale with f decades back to linear. TLOG(p,f) Transforms p defined on a linear scale to a logarithmic scale with f decades. EVNO() Returns the number of the event, e.g. the position in the listmode memory. R denotes the full scale value range. p is typically a parameter name, e.g. SSC or FL1 but can be any expression. f is a integral or decimal number, e.g. 2 or 2.5. FloMax Operating Manual Data Analysis 56

57 Panels - Overview View templates, instrument settings, gating files and report templates have to be defined before they can be selected in panels. These settings are stored in files that may not be removed or deleted. Panels are currently only supported for acquisition. Overview The FloMax panel system allows to predefine complete sets of measurements and run them in an automated way. This approach is often used in immunology to reduce the amount of user interaction. Usually, one complete panel consisting of severeral sample tubes is analysed for one patient sample. Using panels requires the definition of a panel once. Later this panel can be recalled for a panel acquisition or modification. Example panels are supplied in the FloMax Panel folder. The main purpose of using panels is a) to automatically apply labels to the histogram axis, such as "CD3-FITC", b) to automatically save a series of measurements with a common basic name, such as "Smith001" for the first tube, "Smith002" for the next etc. Additionally FloMax panels have unique features to optionally allow automated c) printout of tube result after each measurement, d) selection of histogram templates for each tube, e) selection of special instrument settings for each tube, f) selection of special gating strategies for each tube, g) report generation for the complete multi-tube panel. Tube 1 negative control IgG1-FITC / IgG2a-RPE Tube 2 T-helper cells CD3-FITC / CD4-RPE Tube 3 cytotoxic T-cells CD3-FITC / CD8-RPE Tube 4 NK cells CD3-FITC / CD16-PE Tube 5 T/B cells CD3-FITC / CD19-RPE Tube 6 cleaning tube Example: 2 Color / 5 Tube Panel for a standard immunophenotype analysis. FloMax Operating Manual Data Analysis 57

58 Panels- Definition Define / Edit a Panel 1. Select Panel - Edit Tube Panel... in the menu - the Panel Definition box is displayed. 2. Press Clear Panel to define a new panel. 3. Make sure the parameter names (FSC, SSC, FL1 etc.) are correct. Panel Definition box Typically, view template and instrument settings are the same for all tubes of a panel. Therefore, only select these for the first tubes. Clear these settings for the other tubes - they will then use the settings of the first tube. Gating settings can be included in view templates as well. This is the preferred way since it also stores how gating is applied to the histograms. To remove a tube from the panel, select the tube in the tube list and press Remove Tube. Use a special folder to store the panels, e.g. C:\FloMax\Panels. Define and Add Tube 1. Enter a short description for the first tube, e.g. "neg. control". 2. Enter the tube position (only required if the panel is going to be used with the sample automat). 3. Enter axis labels for the tube, e.g. IgG1-FITC for FL1 and IgG2a-RPE for FL2. These labels will be displayed on the histogram axis together with the parameter names. 4. Check Save for an automated file saving at the end of the acquisition. 5. Check Print for a automated printout of the tube result after acquisition (usually not required). 6. Select a view template for an automated selection of a histogram layout/page (including regions and gates). Clear the view template to use the actual histogram layout. 7. Select instrument settings for an automated selection of special instrument settings for the tube. Clear instrument settings to use the actual instrument settings. 8. Select a gating file for an automated load of gates and regions for the tube. Clear gating to use the actual gating settings (or the gates setup in the view template). 9. Press Add Tube - the tube (with description) now is defined and appears in the tube list. 10. Repeat steps starting from 1. for all other tubes of the panel. 11. For an automated generation of a complete panel report, select a report form and check Create Report. 11. Press OK to finish the Panel definition this panel will stay the active panel (even after program start) until redefined or another panel is load. Save Panel Press Save... - a Save As dialog appears. Specify a folder and filename for the panel and click OK. As extension, ".pnl" is added automatically for panel files. Load Panel Press Load... - a file dialog appears. Specify the folder and filename of the panel and click OK. Panel files use the extension ".pnl". FloMax Operating Manual Data Analysis 58

59 Panels - Optimizing Optimizing Panels - Step by Step I. Optimize the report template 1. Load a basic view template, e.g. "Report.dot" (or another template based on this) in Word. 2. Change the report template according to your requirements (for more details, refer to the Partec Report Manual). 3. Save the view template using a unique name, e.g. "2Color5TubeReport.dot". Report templates must be based on another report template, e.g. "Report.dot" in order to contain the required report macro. Use unique folders for sets of reports, panels, view templates, instrument settings and so on, e.g. "C:\FloMax\Panels". II. Optimize the view template 1. Make sure the view template fits to the report template. 2. Save these view template using a unique name, e.g. "2ColorForm.FCS" Report templates assume histograms and region statistics to appear at a specific location. Therefore, view templates have to be designed carefully in order to fit to the report templates. III. Set up the instrument settings 1. Set up parameter enables, trigger, gain values, lower level -values, while measuring a (typical) sample. 2. Save these instrument settings using a unique name, e.g. "2ColorSetup.ins" IV. Prepare the panel 1. Enter the labels for each tube. 2. Select view template and instrument settings for the first tube only. 3. Make sure the "Save" check is on for each tube. 4. Save the panel file (e.g. "2ColourPanel.pnl"). V. Run the panel acquisition a) Prepare the panel acquisition 1. Select Acquisition - Run Panel Acquisition Enter the sample code (e.g. the patient name "Smith" or barcode by a barcode reader) 3. Select a folder where to save the panel files. 4. Check the panel is correct. Use unique folders for panel results for a series of tube measurements, e.g. "C:\FloMax\Results\Smith". b) acquire tube no. 1 (find lymphocytes) 1. Insert the tube no. 1 for acquisition. It's file will have the name "Smith001.FCS". 2. Correct the gain- and lower level-values if necessary - the file is automatically saved at the end of the acquisition. 3. Select the lymphocyte region. 4. Save the file again by clicking on the save icon (save it together with the region). c) acquire tube no. 2 (negative control) 1. Press the "Next" tube button. 2. Insert tube no. 2. for acquisition 3. Correct the gain- and lower level-values if necessary - the file is automatically saved at the end of the acquisition. 4. Select the quadrant region and apply the lymphocyte region as the FL gate. 5. Save the file again by clicking on the save icon (save it together with the gates). d) acquire other tubes (immune status tubes) 1. Press the "Next" tube button. 2. Insert next tube for acquisition - the file is automatically saved at the end of the acquisition, together with the results. When the last panel tube is finished, the panel report is automatically generated. FloMax Operating Manual Data Analysis 59

60 Generating Reports Copy & Paste Various Ways to Generate Reports - by copy & paste techniques. - by the FloMax report generator (requires Partec Report Module). Copy Button For the copy & paste function, FloMax uses the current screen resolution, which usually provides sufficient quality. For highest quality, you may increase the screen resolution or use the FloMax Zoom function before copying and pasting. Copy & Paste The histogram page can be simply transferred to another Windows application (Word, Excel etc.) by copy & paste: 1. Make sure the histogram page shows all relevant information. 2. Click the Copy Button or Select Edit-Copy in the menu the current histogram page is copied to the Windows clipboard. 3. Activate a document in the other Windows application, e.g. Word or Paint, and select Edit- Paste - the current histogram page is pasted to the open document. A histogram out of a 8-histogram page, copied and pasted into this manual. For most flexibility, it is recommended to copy graphics to a graphics program, e.g. Paint, which is part of Windows. Paint can be invoked by Start- Programs-Accessories-Paint. Paint allows to freely edit the graphics. To copy anything you see on a Windows screen to another application, activate the Window of interest and simultaneously press the Alt and the Print key on the keyboard - the complete actual Window is copied to the Windows clipboard and can then be pasted to other applications, e.g. Paint or Word. A histogram out of a 8-histogram page, copied and pasted after zooming. FloMax Operating Manual Data Analysis 60

61 Generating Reports FloMax Report Together with the Partec Report package (optional), FloMax supports automated generation of reports in Microsoft Word or Excel. Generate Word Report 1. Select Analysis-Create Report in the menu - the report selection dialog appears. 2. Make sure the correct report template is selected, e.g. a Word report document template. If necessary, click on Select button and select another report template. 3. Click OK - Word or Excel is started. The relevant information is transferred to the report. 4. Save or Print the report. For automated reports, the Partec Report software (optional) is required. In the Word report additional information like comments or graphical annotations can be included. Multi-tube reports are currently generated only directly after acquisition of a panel series. Automatically generated Word based on a FloMax report template. FloMax Operating Manual Data Analysis 61

62 Appendix Software Support Maintenance Service Please contact your supplier or Partec for any software requests. FloMax Operating Manual Data Analysis 62

63 Appendix - Mathematical Formulas 1. General Definitions Acquisition By acquisition we understand the experimental measurement process. In an acquisition, from each particle one-by-one, data is collected and stored as one event. An acquisition may cover a large number of N A events, typically between 10,000 and 1,000,000. Parameters For each event, a set of parameters (1...N) can be acquired. Generally, parameters contain information about specific particle properties, e.g. the size of the particle, the DNA content of a cell or the number of bound antibodies to a cell surface. Typical parameters are the pulse heights delivered from different fluorescence channels, when the particles pass the light spot. Other parameters are pulse area or the time parameter, the latter being used for kinetic measurements and denoting when the event was analysed. Events ρ e i = ( x1, i, x2, i,..., xn, i ) denotes a N-parameter event with parameter values of x 1, i, x 2, i,..., x N, i. i = 1... N A is the number of the event. The e ρ i are vectors in a N-dimensional parameter space of event properties. For example, in case of a twoparameter measurement, N = 2 and each event e ρ i = ( xi, 1, xi, 2 ) contains a value pair for parameter 1 and parameter 2. Histograms Histograms are used to classify the events before further analysis is performed. Most frequently, histograms are one- or two-dimensional projections of the complete N-parameter space. Generally, a histogram is composed of channels. Each channel is associated with one or a given combination of parameter values. The channel content specifies, how many of the events with a given combination of properties have been acquired. One-Parameter Histograms A one-parameter histogram contains the distributions of values of parameter values of one of the N event parameters. It consists of N K channels with the channel contents y k, where k = 1...N K specifies the number of the channel. Typically the channel resolution N K is in the range of 128 to 1024 but can be up to in high resolution systems. The histogram channels are numbered from left to right by the index k. Each channel specifies a small value interval [ xk, l, xk, r[ with the left and right limits of the interval x k, l and x k, r, respectively. Ideally, the interval is equal for all channels, i.e. x k, r - x k, l = dx is identical for all k. The channel content equals the number of events with parameter values within the interval x k, l x i < xk, r We will use the index e in order to distinguish histograms containing experimental data from histograms which are mathematically modelled histograms. For the latter, we will use an m, as in y e, k (experimental histogram) (mathematical modeled histogram). y m, k FloMax Operating Manual Data Analysis 63

64 2. Crosstalk Compensation Where it comes from Crosstalk between different optical parameters may generally occur due to several effects: 1) spectral overlap of the dyes (in excitation and/or emission) 2) spectral leakage of the optical filters, e.g. bandpass filters 3) electronic crosstalk. Usually, crosstalk is nearly exclusively due to (1), the overlap of the excitation and/or emission spectra of the dyes. The spectral overlap should be reduced as much as possible by the use of appropriate dyes. Practically, however, reagents are frequently not available with the dye of choice and there is a tendency to analyse more and more dyes at once, which complicates the crosstalk. The influence of (2) is well reduced by selection of appropriate high quality filters. However, if the intensities of the optical signals differ extremely between the parameters, even minuite spectral leakage of the filters (e.g. 1/1,000,000) could make necessary filters to take into account as a cause for crosstalk. In the PAS system, in all circumstances electronic crosstalk (3) can be neglected thanks to completely separated electronics for each parameter. Why to compensate The reason to compensate is to display quantities of the dyes, rather than intensities in a defined spectral range in the histograms. Compensation makes interpretation of data easier. However, it cannot increase the quality of the data itself. How to compensate Assuming a fixed percentage of each parameter contributes to the actually measured signal, the relation between the acquired event x and the quantity of dye x can be written as x = A x The crosstalk matrix A is a N N matrix, which contains all compensation percentages on the non-diagonal elements. N is the number of parameters acquired. x and x are vectors containing N parameter values. Assuming all crosstalk percentages, A is known and the crosstalk can be eliminated by the operation x = A -1 x after the inverted matrix A -1 has been calculated. i R FloMax Operating Manual Data Analysis 64

65 2. Region Statistics The region statistics algorithms work on linear and log-scale data. Generally, log-scale data is transformed into linear data before performing the calculations. The formulas used for the data linearization are described in remark 3. Region Regions define a subset of events. Generally, a region can be defined by a one-parameter region range, a two-parameter polygon, eliptical, rectangular or quadrant region or any other subset. The expression i R denotes a subset of all events, falling into a region R. Regions are identified by a unique name, e.g. RN1, R1, Q1 etc. NR is the number of events belonging to region R. Gate Before evaluating the events belonging to a region, a one- or two-parameter histogram can be gated. The events taken into account in a statistical analysis can be reduced to a subgroup by using gates. These gates can be simply a region, defined in another histogram (or dimension space). Gates can also be defined more complex by a logical combination of several regions defined in different histograms. As an example, a gate can be defined as G1 = R1 OR R2. In a histogram gated by such a gate, all events belonging to region R1 or to region R2 would be included in the region statistics evaluation. To notify which gate was used for the region statistics belonging to a region, this field contains the gate used, e.g. G1, R1, etc., or <None>, if no gate was applied, i.e. all events have been taken into account. Area The Area field is the number N R of gated or ungated events belonging to the region. Area% The Area% field gives the number of gated or ungated events belonging to the region related to all events displayed in the histogram. Remark If a histogram is gated, the Area% value is related to the number of gated events in the histograms. It is not related to the number of all acquired events. Mean-x and Mean-y 1 Mean= x = N R x i i R The mean value (arithmetic mean) can be interpreted as the "center of gravity" in one dimension of the event properties, evaluated on a subset of events, defined by a region R. For a two-parameter region, two mean values can be calculated, one according to the parameter displayed on the horizontal (x or x 1 ), the other displayed on the vertical axis (y or x 2 ). Geometrical Mean. GMean= x = N R = 10 x i i R 1 log 10 x i N R i R Variance Var and Standard Deviation σ Var = σ = ( xi x) N 1 R i R FloMax Operating Manual Data Analysis 65

66 are common statistical entities used in the calculation of the CV% values. In case of two-parameter regions, these values are calculated for each dimension (axis-parameter) independently. Var and σ are meaningfull for any distribution of data, i.e. not only for Gaussian distributions. Coefficient of Variation CV% σ CV % = 100 x is a measure of uniform relative distribution of a subset of events in one dimension. The smaller the CV% values, the more uniform the distribution. By using the standard deviation as defined in the formula further above, no assumption about the distribution of the subpopulation (e.g. Gaussian or not) is made. Such this method, which is in accordance to basic statistics text books, is the most objective way to calculate the CV. FloMax Operating Manual Data Analysis 66

67 Remark 1: Factors Influencing the CVs The CV value(s), analyzed for a subpopulation of particles, is the consequence of several factors. The most important are: I) Sample-dependent influence on CV i) the distribution of the property to be studied ii) the preparation process which affects the transformation of the property (i) into a physically measurable signal (e.g. scatter light, fluorescence) iii) the statistical properties of light being emitted from each particle (photon statistics) II) Instrument-dependent influence on CV iv) the detection efficiency of the optical system and photodetectors together with the related photon statistics v) the precision of the fluidics/optics system which affetcs the uniform illumination of particles vi) the stability of the light source vii) viii) noise of electronics amplifiers (e.g. logarithmic amplifier) errors introduced by limited resolution of digitalization and data handling (by analog-todigital-converters and software). These factors can be classified into I) sample dependent, as i), ii) and ii), and II) instrument dependent, as iv) - viii). Due to the many factors affecting the CV, in case of an unexpected high CV, these factors have to be examined carefukky in order to separate between sample and instrument depending factors. FloMax Operating Manual Data Analysis 67

68 Remark 2: Alternative Methods to Calculate CVs Generally, different ways of calculating CV% values are used in the analysis of flow cytometry data. Based on the assumption that the peak-data is distributed according to a Gaussian, all methods should deliver the same value. However, since experimental data in general is not strictly distributed according to a Gaussian, different methods can result in different values, which cannot be directly compared to each other. An objective method not assuming any kind of distribution but taking the data as-is is the one described further above. As an example, a method to calculate the CV% directly from the peak width, differing from the method above and assuming a Gaussian distribution is as follows. This method can easily be used for a fairly good estimation of the CV "by looking" on the histograms. The "Peak Width Method" For a Gaussian distribution y exp ( x x ) 2 0 = 2 2σ with a maximum of 1, the peak width at a height of y 2σ = exp( -0.5 ) = 1 / e = simply is δx = 2. σ. Thus, assuming the distribution of the peak is Gaussian, the formula σ CV % = 100% x 0 can be used with σ evaluated from the peak width by using σ = δx / 2.. Consequently, it is sufficient to evaluate the peak width δx (in terms of channels) at a relative height of y 2σ and relate this width to the position x 0 of the channel with the highest content, as in the simple formula δx CV % = 100% 2 x 0 The CV calculated by this method most frequently delivers smaller CV values. This is because for experimental data, the peak distributions tend to be relatively less uniform at the bottom of a peak compared to an ideal Gaussian, whereas the peak width is evaluated more at the top. Example: Application of the "Peak Width Method" Assume a peak shows up at channel 100 with a width of 2 channels at 0.6 of its height. The CV can then easily be calculated to be 2 CV % = 100% = 1% Attention: CVs and log-scales Note that this method should not be applied for data displayed on a log-scale. On a log-scale, for example, ideally, the absolute width of a peak is independant of its position. Consequently, the appearent CV, i.e. the peak width related to its position, would significantly depend on the position. A CV calculated this way is meaningless. To calculate a significant CV, data on a log-scale first is to be transformed to a linear scale, as described in remark 3. FloMax Operating Manual Data Analysis 68

69 Remark 3: Statistics on log-scale Data For the calculation of CVs, Mean, and other statistical values in histograms with logarithmic scale, data needs to be linearized before a meaningfull calculation can be performed. For the linear, 3- and 4-decade logarithmic amplification, the photodetector signal voltages u i (0... u max, e.g. u max = 10V) corresponding to each event and parameter are transformed to event x i values (0... x max = 65,535) according to x i = x. max u i / u max (linear) x i = x. max (1/3). log{ u i / (u max / 1,000 )} (lg3: 3 decade log) x i = x. max (1/4). log{ u i / (u max / 10,000 )} (lg4: 4 decade log) To linearize the x i values back to the original signal voltages, the following reverse formulas are applied: u i = u. max x i / x max (linear) i xi 3 xmax max /1, u = u (lg3: 3 decade log) i xi 4 xmax max /10, u = u (lg4: 4 decade log) Note that in some systems, this transformation back to the linear scale is not performed. Instead, sometimes the calculations are performed as if data was measured linear and then transposed to a linear channel. This method is sometimes called "Lin-Log" mode. As an example, in this mode, the mean value is calculated as 1 x i NR i R Cd Mean= x = 10, where C d is the number of decades on the log-axis. It is not recommended to use this method, since results depend on the position of the peaks, the number of decades on the log-scale, etc. and thus cannot be directly compared among different measurements or systems in a straight forward way. FloMax Operating Manual Data Analysis 69

70 3. Numeric Peak Analysis The numeric peak analysis finds peaks in a histogram (step 1) and fits a mathematical model to the experimental histogram (step 2). Note that this algorithm assumes the histogram scale to be linear, i.e. no linearization is performed. The peak searching step invoked before the fit uses smoothing and minina / maxima finding algorithms. This operation finds the number of peaks N P and their positions. In the fitting step, each of N P peaks found in a one-parameter histogram is mathematically modelled by the Gaussian 2 ( x x ) j,0 y j = M j exp 2 2σ j with each 3 parameters, 1) the height M j, 2) the position x j, 0 and 3) the width σ j of the peak. j = 1... N P denotes the number of the peaks, numbered from left to right in the histogram. Note that for a Gaussian distribution, the mean value and the position are identical due to its symmetry, x j = x j,0. The complete histogram is modelled to be a sum of such Gaussian distributions: N = P y m y j. j= 1 By a recursive method, the numeric peak analysis optimizes all of the 3. N P parameters M j, x j, σ j such, that the 2 deviation of the mathematical model to the measured histogram data χ is minimized. The peak analysis results are taken from the mathematical model (with the optimized parameters), according to the following list Peak The number j of the peak found in the histogram, from left to right. Index Index = x j, 0 / x1,0 is a relative measure of the mean position of peak j, related to the leftmost (first) peak in a histogram or another reference channel position, entered manually. Mean Mean = x j = x j,0 specifies the position of peak j, in channel coordinates. Area Area = y j ( x) dx is the area of the Gaussian distribution. This integral can be evaluated by a numerical integration or use of the error function erf. Area% Area % = Area / N A 100 is the percentage of the area of the Gaussian distribution related to all events in the histogram. CV% σ CV % = 100 x is a relative measure of the width of the Gaussian peak, related to its position. FloMax Operating Manual Data Analysis 70

71 ChiSqu. ChiSqu. = χ = N 2 1 N K K k= 1 ( ) 2 y y m, k e, k is a measure for the convergence of the numerical fit, i.e. how good the mathematical model y m, k fits to the experimental data y e, k. Smaller values correspond to better fits. FloMax Operating Manual Data Analysis 71

72 4. Numeric Cell Cycle Analysis Note that this algorithm assumes the histogram scale to be linear, i.e. no linearization is performed. The numeric cell cycle analysis operates in two steps. In step 1, the G1 and G2/M peaks are identified in a oneparameter DNA histogram by using historam smoothing and minima / maxima finding algorithms. In step 2, the G1 and G2/M peaks are mathematically modelled by the Gaussian distributions 2 ( x x ) j,0 y j = M j exp 2 2σ j with each 3 parameters each, 1) the height M j, 2) the position x j, 0 and 3) the width σ j. j = G1, G2M identifies the peaks. Remark In case an aneuploid peak is included in the fit, it is as well defined by the above Gaussian formula, i.e. j = ANEUPLOID. The S-phase in the DNA histogram is modelled as a constant M S in the region not overlapping with the G1 and G2/M phase and an overlap function erf in the overlap regions according to 2 ( x x ) ( x x ) 2 G1,0 G2M,0 ys = M S erf erf σ G1 2σ G2M G1 and G2/M belong to the same cycle as the S-phase. erf is the statistical "error function" (an integrated Gaussian). The complete cell cycle is mathematically modelled to be a sum of such Gaussian distributions plus the distribution of the S-phase: y = y j + ys. j Remark In case of a fit to two cell cycles, two of the above distributions are summed up, each with individual parameters. By a recursive method, the numeric cell cycle analysis optimizes the parameters M j, x j, σ j and M S of one (or two) cycles such, that the deviation of the mathematical model to the measured histogram data 2 χ is minimized. The cell cycle analysis results are taken from the mathematical model (with the optimized parameters), according to the following list G1% S% G2M% ( G1 + S + G2M ) 100 G 1% G1 / = Area Area Area Area ( G1 + S + G2M ) 100 S % S / = Area Area Area Area ( G1 + S + G2M ) 100 = Area Area Area Area G 2M % G2M / Herein, G1 Area, S Area and G2/M Area are the areas of the distributions, corresponding to the G1, S and G2/M phases of the cell cycle. Remark As can be seen from above, the sum of these percentages sums up to 100%, G1% + S% + G2/M% = 100%. FloMax Operating Manual Data Analysis 72

73 CV%G1 and CV%G2 σ j CV % j = 100 x j is a relative measure of the width of the Gaussian distribution modelling the G1 and G2 peak, respectively, related to their position. Mn G2M and Mn G1 x j = x j,0 specifies the mean values or the position of the G1 or G2/M peak, respectively, in channel coordinates. G2M/G1 G2M/G1= Mn G2M / Mn G1 is the ratio of the positions of the Gaussian peaks, corresponding to the G2/M and G1 phases of the cell cycle. Generally, this value can be used as a check if the peaks identified to correspond to G1 and G2M indeed belong to the same cycle. Generally, this value must be very close to 2. ChiSqu. N k 1 2 χ = N k k= 1 ( ) 2 y y m, k e, k is a measure for the convergence of the numerical fit, i.e. how good the mathematical model fits to the experimental data. Smaller values correspond to better fits. FloMax Operating Manual Data Analysis 73

74 Appendix Computer Recommendations Computer for FloMax data analysis (offline) Operating System Microsoft Windows 95/98/NT/2000* Pro/Me* Processor > Pentium 300 MHz RAM > 64 MB Harddisk > 2 GB Graphics > 1024 x 768 pixels CD-Drive CD-RW recommended Floppy Disk Drive 3.5, 1.44 MB Printer Inkjet or Laser printer *Please inquire for support of instrument control under Windows NT/2000. Printers (printers tested with FloMax) Canon Bubble-Jet BJC600, HP DeskJet 690C, 815C, 850C HP Laser Jet III Any other printer supported by Windows might be suited as well. Please inquire for actual information about tested printers. FloMax Operating Manual Data Analysis 74

75 Appendix Software Specifications FloMax Software Specifications FloMax software package for multiparametric flow cytometry data analysis for data in FCS 2.0 standard format from any flow cytometer and for instrument control and data acquisition in conjunction with Partec flow cytometers. Delivery Form CD with software and sample files. Operating Manual. Laboratory license for several computers in one laboratory. General Windows 95/98/ bit software. Support of long filenames according to Windows standard. Support of up to / 16 bit channels data resolution. Data resolution in graphs: channels for 1-parameter histograms, 32x x1024 for 2- parameter dotplots. Calculations based on double precision arithmetics (64bit). FCS listmode storage according to Data File Standard for flow cytometry published by the Data File Standard Committee of the Society of Analytical Cytology (Cytometry 11: p , 1990). Support of dual screen computers. Support of network connection to Macintosh computers (with optional software). Analysis True Volumetric Absolute Counting (for data from Partec flow cytometers). Determination of cell concentrations by volumetric absolute counting. Concentrations of subpopulations (cells / ml) for each region or gate. Reselection of subpopulations for absolute counting. Gating System Multiparameter gating: Up to 32 parameters. Regions and quadrants: Up to 32 one- and two parameter ranges / regions / quadrants. Gates: Up to 32 logical combinations of regions. MultiColor Gating. PriorityColor Gating. Peak and region statistics. Movement and resizing of regions and quadrants. Storage of gated data for export to other analysis software. Storage of gating protocols. Colour Crosstalk Compensation System N-colour crosstalk compensation system with autofluorescence entry. Graphical or numeric entry of compensation values. Stepwise compensation for 3-4 colour analysis. Recovery and reanalysis of uncompensated original data. Storage of compensated data for export to other analysis software. Storage of compensation settings. Calculated Parameters Addition of calculated parameters according to user-defined arithmetic expressions or predefined functions. Scaling of data. Nonlinear transformations for better visualization. Ratio analysis for any combination of parameters. Linear-logarithmic / logarithmic linear transformation with any no. of decades by software. Pseudo-time parameter by event number function. FloMax Operating Manual Data Analysis 75

76 Analysis and Automated Results Peak- and Cluster Analysis and statistics. Numeric DNA-cell cycle analysis with different fit models. Numeric ploidy analysis with Gaussian fits. Stability check by rerun from file function. Barcode reader support. Multi-page - multi-tube report generation in Word and Excel. Easy to adapt Microsoft Word/Excel report templates and macros. Quality check report examples. High quality colour or b/w reports on all Windows printers. Full network support via Windows operating system. Instrument Control and Realtime Acquisition Flow Cytometer (only for Partec flow cytometers). Realtime display in up to 8 one- / two-parameter histograms / dotplots per analysis. Realtime N-colour crosstalk software compensation system. Crosstalk compensation for any combination of parameters. Recovery of original data and readjustment of compensation settings any time without need to rerun the sample. Acquisition of more than 10,000,000 events. Acquisition event rate >10,000 events/s. Preset of particle number and/or time of analysis. Scatterplots FSC-SSC. Standard 1-, 2-, 3-, and 4-colour analysis. 2-wavelength acquisition/delay system (laser/laser or laser/uv-lamp), tolerant for pressure changes due to time window technique. Support for 8 optical channels. Doublet discrimination system: Peak height, area and width acquisition; 3 parameters for each channel. Acquisition of time parameter for kinetic analysis. Rescalable time axis time = 25 parameters. Flexible trigger selections: Leading trigger and trigger on all parameters. High speed realtime cell sorting control based on any gate (free region shape). 16 bit data resolution for each parameter (65536 channels) for improved precision of numeric analysis. Automated BioSafety fluidic control. Absolute counts of subpopulations (cells / ml) for each region. Offline reselection of subpopulations for absolute counting. Automatic file storage and printout. Panel system allows flexible automated multi-tube data acquisition. Up to 100 tubes per panel. Storage of different panels. Histogram layout templates. Instrument setup protocols for standard applications. Sample Autoloader Control (for Robby sample preparation system) Automated sample analysis for up to 36 tubes per caroussel. Works together with panel system. Possible manual selection of tubes in single tube mode. FloMax Operating Manual Data Analysis 76

77 Appendix Menu Function Overview File New Open Open Next File Open Prev File Close Save Save As Save Options Print Now Print Print Preview Print Setup Exit Panel Edit Tube Panel Edit Copy Page Acquisition 1 Instrument Settings Run Run Tube Panel Run Automate Previous Tube Next Tube Start Pause End Clear Histograms Clean Sheath Fluid Prime View Quick Zoom Zoom + Zoom Page Layout Color Gating Region Statistics Gate Statistics Instrument Settings Toolbar Status Bar Instrument Controls Window Cascade Tile Arrange Icons Close All Help About FloMax Analysis Rerun From File Undo Analysis Compensate Crosstalk Gating Regions - Polygon Regions - Quadrant Regions - Region Range Gating Region Statistics Cell Cycle Analysis Peak Analysis Calculate Parameters - Ratio - Formula Create Report 2 1 only available if Partec instrument is connected 2 requires optional Partec Report software FloMax Operating Manual Data Analysis 77

78 Index Absolute Cell Counting, 19 absolute counting, 40 area, 52 ASCII, 41, 54 background, 10 Barcode, 24 calculated parameters, 55, 56 defining, 56 examples, 57 function overview, 59 lin-log transformation, 58 scaling, 57 cell cycle analysis, 17 fit G1/G2 only, 51 fit one cycle, 51 numeric, 73 Cell Cycle Analysis, 51 channel count, 31 information, 31 position, 31 channel content, 31 channel values, 56 chi square, 51, 52 click, 4 cluster, 12 Cluster Size, 49 coefficient of variation, 67 color gating, 30, 43, 44 compensation, 65 exporting, 48 notes, 49 save, load, 47 undo, 46 Concentration. See Absolute Counting copy & paste, 60 count, 40 count/ml, 40 counting volume, 27 crosstalk compensation, 65. See compensation setup, 46 cv, 40, 52. See also coefficient of variation CV, 51 DAPI, 11 DNA Cell Cycle Analysis, 17 Dotplot, 12 doubleclick, 4 equation editor, 56 Excel, 16 Exit, 14 formulas, 64, 75, 76, 78, 79 G1, 51 G2/M, 51 gate, 30, 40 gates moving, resizing, deleting, 39 gating applying gates, 36 defining regions, 35 exporting, 42 logical gates, 37 multicolor, 43 saving and loading, 38 statistics, 40 geometrical mean, 40 HBO, 11 histogram export, 54 properties, 28 resolution, 29 Histograms, 10 Immunophenotype Analysis, 15 index, 52 Installation, 7 intensity, 31 Kinetic Analysis, 16 label, 29 l-l. See lower level logical gates, 43 lower level, 10 lymphocytes, 11, 44 mathematical formulas, 64, 75, 76, 78, 79 mean, 40, 52 arithmetic, 66 geometric, 66 Microorganism Analysis, 18 noise, 10 Open, 23 Page Layout, 27 paint, 60 Panels Definition, 21 Optimizing, 22 Overview, 20 parameter, 10 peak analysis batch mode, 53 numeric, 71 numerical fit, 52 range method, 53 percentage, 52 percentages, 40 polygon region, 35 preselected events, 27 printer recommendations, 75 Printing, 25 pseudo time, 56, 57 quadrant, 35 random bias, 46 Random Bias, 50 range, 27, 35 FloMax Operating Manual Data Analysis 78

79 ratio analysis, 55 region, 40 registration, 8 Registration, 8 reports automatical, 61 copy & paste, 60 rerun, 32 right click, 4 S, 51 Save, 24 Save As, 24 Save Options, 42, 48 Screen Elements, 9 specifications, 63 spectral analysis, 55, 56 spectral overlap, 45 standard deviation, 66 Start, 14 statistics, 40, 66 log scale data, 70 saving to file, 41 support, 62 tab text, 54 tab-text, 41 time, 56, 57 transformations lin-log, 56 nonlinear, 56 ungated, 40 zoom, 34 FloMax Operating Manual Data Analysis 79