Description of Old GELLAB-II 2-Dimensional Electrophoretic Gel Data Mining System

Description of Old GELLAB-II 2-Dimensional Electrophoretic Gel
Exploratory Data Analysis System

Last GELLAB-II Unix Version: 1.3.52 - Released August 7, 1993

GELLAB-II is an integrated collection of Unix programs for the analysis of 2-dimensional polyacrylamide electrophoretic gels (2D PAGE). The last available version was 1.3.52-Release (1993). It was written in C and used X-Windows for interactive graphics as well as Tektronix graphics for plotting. It was developed by the Image Processing Section, Laboratory of Experimental and Computational Biology, Center for Cancer Research in the National Cancer Institute.

The GELLAB-II programs were documented in a reference manual describing GELLAB-II [31]. The reference manual also has tutorial examples of running GELLAB-II using a sample set of gels supplied with the system. A poster is available that illustrates some of the aspects of GELLAB-II.

The following describes GELLAB-II as it existed in 1993. It has not been maintained since then and parts of it are being be re-released as open source as we refactor parts of it into the Open2Dprot project on SourceForge.Net at http://open2dprot.sourceforge.net/. We are encouraging the integration of other 2D gel analysis and related proteomics software (LC- MS etc.) into Open2Dprot as well. The following is taken from the 1993 description of GELLAB-II.

Top-level description of GELLAB-II

Figure 1 and Figure 2 illustrate the data reduction processing performed in the original GELLAB-II analysis.

Figure 1. Block diagram of the original 2D-gel analysis GELLAB-II system. Programs associated with major steps of GELLAB-II are indicated in "[...]". Gel images are acquired by scanning with a camera interfaced to the UNIX system and saved on the computer disk in step 1. Accession information about the set of gels is also used to update an accession file. Landmark spots are then manually selected that are well defined spots spaced fairly evenly throughout the gel - with more landmarks in regions with higher distortion in step 2. Using gel image flicker alignment, the landmark spots are aligned for all of the gels with a Representative gel (Rgel). The gel images are then segmented and measurements made of the spots that are found in step 3. This information and the raw segmentation data is then used to pair corresponding spots in the remaining gels with the Rgel in step 4. The set of gel pairings with the same Rgel may be merged together to form a list of sets of equivalent Rspots called the composite gel database (CGL) in step 5. Thus a Rspot set (most likely) contains corresponding spots from all the gels in that it occurs. Finally, in step 6 data mining is performed on the PCG database, reports are created, and Rmap and mosaic images of statistically significant spots can be generated and displayed. These results can be annotated and analyzed further using additional software.

Running the programs

Programs are invoked in GELLAB-II through the Unix command-line by naming the program and specifying optional arguments. Commands may be combined in Unix script files to implement a batch mode. All programs have several common Unix-style command-line switches to facilitate learning the consistent user interface. This subset of switches is useful in learning how to run particular programs. Descriptions of the algorithms used in these programs are given in many of the GELLAB-I papers listed in the References [4,7-13,17,20,28-30]. [The original GELLAB system (GELLAB-I) was written in the SAIL (Stanford Artificial Intelligence) Language - an Algol dialect. It was then translated using the PSAIL compiler [23,26] to C and then integrated with X-windows and Unix to create GELLAB-II.] Good introductory papers that describe the basic GELLAB analysis are [4] and [10] with [12] being a more general and detailed summary. Extensions to the early system are discussed in [11,13,17-18,20,28-30]. A comparison of aspects of 2D gel database analysis systems is given in [28]. This document lists the GELLAB-II programs and describes the basic steps in performing a 2D gel analysis. Full documentation of GELLAB-II is given in the Reference Manual [31]. A few selected examples of running some of these programs (Section 5) are included at the end of this document. Each GELLAB-II program can provide help on how it should be run using the following command line switch options.

-info
Print detailed information about the program including what it does, how to run it, and literature references specific to that program.
E.g., cgelp2 -info.

You can do use this with the Unix more(1) program to page through the documentation slowly. E.g., cgelp2 -info | more
or use it to create a printable file. E.g., cgelp2 -info > cgelp2.info.

-version
Print the version number of the program. E.g., cgelp2 -version

-usage
To print a quick list of command line switch usage for the program. Note that you need only type that initial part of a program's switches that makes them unique. This is indicated by that part of the switch being upper case in the usage print out.
E.g., cgelp2 -usage or cgelp2 -usage | more

The X-windows System window manager program (under X11R4) twm(1) with startup file .twmrc creates three menus: TWM-WINDOWS (left button), WINDOW-OPS (middle button), and APPLICATIONS (right button). Dragging the mouse to the GELLAB-II selection in the APPLICATIONS menu, causes a GELLAB-II programs menu to pop out to the right when the mouse is in the right side of the menu. These main GELLAB-II selections are listed below. Continue dragging the mouse to the right part of the selection in the menu desired. A function-specific menu now pops out to the right. At this point, select the specific sub-function you want and release the mouse button. Upon invoking a program from the GELLAB Tasks menu, it prompts with a popup form-dialogue window for additional information required for the command line. Press HELP in the dialogue window to list the (Emacs-like) text editing commands. After you have answered the questions, press OK to perform the function or CANCEL to abort it. When the function starts, it pops up a X11 xterm window that disappears when the function being performed is finished.

GELLAB Tasks

Accession gels into the gel accession database file for subsequent processing.

Autopair gels (GSF data) to Gel Comparison File (GCF) paired-spot lists.

Convert image files to PPX images (PPX is the GELLAB standard image format).

Convert PPX images to PostScript for subsequent display or plotting.

Debug PPX image file (view numeric data at the pixel level).

Display gel images or derive image map images.

Draw Rmap (Representative gel with spots overlaid) as a GSF plot.

Edit Gel Comparison File (GCF) paired-spot lists.

Edit Gel Segmentation File (GSF) spot-lists.

Gel Database Manager for a Composite Gel Database (CGL).

Landmark gels by interactively defining a small number of common landmark spots.

Make GELLAB batch scripts to accession, landmark, segment spot, pair spots, and build initial composite gel database.

Mosaic to create derived PPX images (a montage) of local gel regions surrounding a selected spot across a set of gels.

Pair gels to a Gel Comparison File (GCF) between a gel and the Representative gel and generate derived image files showing pairing.

Print GELLAB-II gel.rc database project resource file in user-friendly form.

Rmap image generation of an overlay map of a subset of spots of interest on a gel PPX image file.

Segment a gel image to a Gel Segmentation File (GSF) and derived PPX image files.

Xpix image display to display one or more PPX image files.

1. Brief Descriptions of GELLAB-II Programs

Most GELLAB-II programs generally require one or more arguments so you should read the individual programs' documentation (using -info as suggested above) prior to attempting to run them. GELLAB-II programs use a database project resource file called gel.rc in the user's current path to provide state information. This includes various directories for image files, gel database files, and other intermediate files (see 4. Unix Directories and Support Files). If you do not have this file in your path, running any GELLAB program (such as pgelrc that prints a user-friendly form of the project resource file gel.rc) can be used to prompt the user in defining the initial gel.rc file. A batch script generation facility, that is part of GELLAB (makjob), can greatly automate running these programs. Then minimum investigator intervention is required for major parts of its operation in composite gel database preparation.

The GELLAB-II programs are listed below. Corresponding literature references are specified by [...] listed in the section Reference Manual [31].

The GELLAB-II programs include:

accppx - display gel image(s) given their accession number(s)

autopair - pair 2 gels (GSF spot lists) by automatic pairing to create GCF file [this program was not released as of the last GELLAB-II release].

cgelp2 - interactive Paged Composite Gel database analysis system with both command line (also used for batch) or graphical user interface.

cmpgl2 - paired 2 gels (GSF spot lists) using landmark database to create GCF file.

dendrogram - hierarchically cluster protein spots or gel samples and draw a plot.

dwrmap - draw Rmap numbered plot of GSF spot list.

getacc - multiple gel accession data to gel accession database: images, information, calibration.

landmark - interactive graphics acquisition to enter paired gel landmarks into the gel LandMark Set database.

makjob - create GELLAB-II scripts for batch processing gels to accession gels, landmark gels, segment gels, pair gels with respect to the reference gel and build the initial database.

markgel - generate Rmap image of a set of spots from a cgelp2 generated (.sps) data file.

mosaic - generate mosaic (montage of gel panels of a local spot region from multiple gels) image from cgelp2 generated (.sps) file.

pgelrc - "pretty print" the gel.rc GELLAB-II database project resource state file.

plotn - plot GELLAB-II Universal Graphics Files (.ugf) to Tektronix or postscript graphics.

ppxcvt - convert foreign image formats to Portable PiXture file used by GELLAB-II.

ppxodt - Portable PiXture (.ppx) file image debugger.

ppx2ps - convert PPX image file to Postscript.

sg2gii - segment gel image to Gel Segmentation File (.gsf).

tek2psG - convert tektronix 4010 graphics input to PostScript (derived from E. Moy's tek2ps program).

Xpix2 - with X11R4 display and manipulate PPX images.

Xpix11 - with X11R4 display and manipulate PPX images (later version).

2. Performing An Analysis On A Set Of 2D Gels From An Experiment

A typical GELLAB-II analysis of a set of 2D gels is a data reduction process. It analyzes a set of gels of the same material but with different experimental conditions to produce lists of spots with similar specific attributes. These subsets of spots are clustered using statistical techniques. The major steps of an analysis are: (1) accessioning gel experiments and scanned image files, (2) spot quantification, (3) gel pairing, and (4) composite gel database construction, and searching and display of search results of different views of the database. The data reduction is shown in the sets of files in the next figure.


a) {Initial image files Gi.ppx} and {Accession file .id}
     G1              G2                Gn
     |               |  ...            | spot segmentation/quantification
     v               v                 v
   GSF = {spot list} GSF = {spot list} GSF = {spot list}
      1                 2                 n
    {Gel Segmentation Files (GSF) .gsf}


b) GSF        GSF      LMS (n-1) landmarking with Rgel GSF
      r          i        ri                              r
    |          |        |        {LMS are in landmark DB file .lm}
    v          v        v
    ----------------------       (n-1) gel comparisons with GSF
             |                                             r
             v
            GCF = {{spot pairs} , {spot pairs} ,..., {spot pairs} }
               i               A              B                  K
            {Gel Comparison Files (GCF) .gcf}


c) {GCF , GCF , ..., GCF }
       1     2          n-1
      |
      |                          cgelp2 composite gel DB construction
      v                          [CGL is stored in (PCG) .pcg file]
     CGL=({Rspot }, {Rspot }, ..., {Rspot }).
                1 |       2              z
                  v
     --------------------------------------------------
      |           |         |            |           | Derived export files
      v           v         v            v           v
    {DB .cgl} {SPSS .sps} {Table .tbl} {Plot .ugf} {Inquire .inq} ... etc.
                |                        |
                v                        v
          {Derived image files .ppx}   {Line drawing plots}

Figure 2. Some of the files used in a GELLAB-II gel analysis. Data file structures and corresponding file extensions used in the gel analysis. A GELLAB-II file extension is a 2 or 3 character name preceded by a ".", (e.g. ts3pcg.cgl). a) Gel Segmentation Files (GSFs) are produced by segmentation of the gel images by sg2gii. b) Gel Comparison Files (GCFs) are produced by comparing GSFs using landmark spots with cmpgl2 or autopair. c) The Paged Composite Gel (PCG) database is constructed by merging GCFs with cgelp2. The PCG DB is a 3D data reduction of the original set of gel images and accession information. The PCG DB is the realization of the Composite GeL database (CGL) model. The SPSS (.sps) file is an exportable file suitable for input to SPSS or other statistics packages. The markgel and mosaic programs uses the exported .sps files. The derived export files are other types of data derived from the PCG DB (see Figure 1).

Briefly, this data reduction is achieved as follows:

(1) To create a new GELLAB database, one enters a set of gel data into the database. This process is called accessioning. Accession a set of scanned gels by entering the information about them with the getacc program that: (a) assigns an accession number to each gel; (b) converts, if necessary, scanned gels into Portable PiXture (PPX with .ppx file extension) files; (c) requires the experimenter to enter associated experimental study accession information; and (d) if needed, calibrates the optional ND step wedge scanned with the gel and defines an active region in the gel image called the computing window. The information from (c) and (d) is entered into an accession file. The PPX image header contains information on OD calibration, image size, computing window, etc. and is described in the Reference Manual. An accession file typically has a gel prefix, a 3 character project name, and a .id file extension. E.g., gelts3.id. During a data acquisition session, one would enter a number gels and at the end of the session getacc would prompt you for a few pieces of information necessary for further processing. These include (a) the name of the Reference gel or Rgel, (b) a three character project prefix used for all files associated with the project, and (c) the names of the different experimental classes to that the different gels belong. It then invokes the makjob program to generate Unix batch scripts to interactively landmark these gels, segment or extract list of quantified spots from the gels, pair N-1 of the N gels with the selected Reference gel, and construct the composite gel database and perform some initial statistical tests. The makjob program also lets you directly generate these batch scripts for different sets or subsets of gels that have been previously accessioned.

(2) Spot-list extraction and quantification is performed by the sg2gii program that results in a Gel Segmentation File (.gsf) and an optional extracted spot image file. The GSF file contains position and quantitation information for all spots in a single gel and must be further processed to compare it with other gels.

(3) Pairing of GSF spot lists from two gels (one of that is the Rgel) is performed by the cmpgl2 program. The output is called a Gel Comparison File (GCF) and is a file with a .gcf file extension. It consists of the pairing of spot data from the two GSF input files. The pairing program also requires a list of a small number of corresponding landmark spots for the two gels being paired. This is stored in the LandMark Set (LMS) data base file that typically has a lms prefix and .lm file extension. LMS data can be acquired several ways: (a) using the landmark interactive graphics program running under X-windows, (b) using program dwrmap to draw Rmaps from GSF data that can be plotted with plotn. The landmark numbers can be read manually from the plots and then entered via a terminal session using the landmark program. A third method involves using the Xpix program in its "compare" mode to interactively generate the landmark coordinate pairs that can then be text edited into the proper LMS DB format. Needless to say, use of the landmark program is encouraged as it is by far the easiest and most accurate.

(4) Construction of the Paged Composite Gel DB (PCG DB) is performed by cgelp2. Cgelp2 requires a set of N-1 GCF files for N gels since the Rgel is included in each GCF. By "paged" we mean that as the data base is too large to fit in memory, pieces of it are paged in and out of memory from the actual PCG DB disk file (that has a .pcg file extension). [Database can be easily over 100 Mbytes for databases consisting of more than 100 gels.] The accession file is also accessed to extract the "study" information for each gel in the PCG DB. This information is used for automatic classification of gels into the current experimental classes of gels and other analyses. Exploratory data analysis really starts once the PCG DB is constructed. The particular strategy to follow is outside the scope of this document but is discussed in many of the papers listed in the references on cgelp2 and papers on particular biological problems.

The initial batch script may be generated by getacc or makjob: (a) constructs the PCG DB file, (b) constructs an initial experimental gel classification based on accession file study information, (c) normalizes the protein concentration values between gels using the Ratio-List method and reorders spots in all Rspots sets in the PCG DB based on this normalization. It (d) then performs an initial F-test and t-test at p-values of 0.90, 0.95, and 0.99 for all of the experimental classes. It performs a Wilcoxon-rank-sum test of classes 1 and 2 at 0.90, 0.95, and 0.99 for classes 1 and 2 as well as a missing-class test. It also computes and displays histograms of various Rspot set spot features for the entire PCG DB to aid in setting the initial prefilter parameters. Subsequent analyses consist of changing the view of the PCG DB, performing searches in the new view, and displaying this transformed data as images, plots, tables, lists, etc. See [12-13,17,28-30] for more discussion or using GELLAB for exploratory data analysis.

3. Details on Programs

accppx
display one or two gel image files using Xpix given the accession number(s) and optional picture prefix type. Optional picture types include: "l" for landmark Rmap, "m" for Rmap images, "y" and "z" for segmented spot images produced by sg2gii, "c" for segmented connected component images, "w" for the set of montage images generated by the mosaic program. (See the PPX file definition in the glossary as well as sg2gii and markgel for more information). Also search accession file and disk for status of gels and PPX files on the system. If a picture name is used instead of a picture number it tries to find that file, adding a .ppx file extension if needed. The current gel.rc resource file specified picture paths are searched first before searching the user's PATH environment variable.

annotate
is an interactive spot annotation system. It reads the spot lists for the specified gel that is displayed in a window. Using the mouse, spots of interest may be queried for their annotation or be specified to add or edit their annotation. Various other annotation specific functions are available including generating derived Rmap images with investigator specified annotation for selected spots.

autopair [EXPERIMENTAL]
is an automatic gel pairing program that can be used to replace cmpgl2 and landmark programs for doing spot pairing. It requires a minimum of one landmark point in each gel to get started and should be more accurate as well. Pairing results are better with similar gels. It does take somewhat longer however to compute the spot pairings.

cgelp2
runs the Paged Composite Gel database analysis system. This builds a PCG DB file from a set of Gel Comparison Files (GCF)s produced by the cmpgl2 program. When running cgelp2, additional information is available on top level commands by typing HELP to list all of the top level commands or HELP specific-command. For example, type HELP HELP to get more information on the HELP command. There is also an ?APROPOS facility for finding relevant commands. [4,9-10,12-13,17-18, 20-22,28-30]

cmpgl2
runs the gel pairing program, that generates a Gel Comparison File (GCF) (.gcf) from two Gel Segmentation Files (GSF) (.gsf) produced by the sg2gii segmentation program. It also requies a landmark set DB (LMS) entry for the two gels being paired. [4,8,10,12]

dendrogram
generates a dendrogram cluster analysis plot. It uses cgelp2 produced SPSS-compatible (.sps) or INQUIRE (.inq) text files. It can cluster a set of Rspots as a function of density of a set of gels or cluster a set of gels as a function of the density profile of a set of Rspots sets. It also plots the results after they are generated and/or makes an optional .ugf plot file. A data file (.dgm) is also produced that contains numeric cluster analysis information. [24]

dwrmap
given a Gel Segmentation File (GSF) Plots a Rmap from the .gsf file. It plots a Rmap with spots labeled by their GSF spot number. This can be used with the landmark program to manually generate the landmark set data entry. It also plots the results after they are generated and/or makes an optional .ugf plot file. [13]

getacc
in a data acquisition session, acquire 2D gel images and their related accession information that is appended to the accession file. At the end of the session, ask a few questions regarding the type of experiment to be performed. Then generate Unix batch scripts to a) interactively landmark the set of gels, b) segment the gel images into GSF spot lists, c) pair GSF spot lists into GCF paired spot lists, d) merge the GCF files by constructing a PCG DB file and e) perform an initial statistical analysis of the PCG DB. Using the -edit -ask: switches you can accession gel images scanned elsewhere. You may also edit entries in the accession file. It allows you to alter different fields for accession entries in the accession file. Image size defaults to 512x512 but any size (e.g., 1024x1024) can be specified. [4,7,12]

landmark
is an interactive X-windows graphics program to landmark two gels. This process defines a small set of corresponding spots (10 to 20) in each of the two gels. These spot positions are used to update an entry in the LandMark Set (LMS) data base file that is used by other programs including the spot pairing program cmpgl2. It can also use a previously defined LMS entry to indicate where the landmarks are in the Rgel image when landmarking another gel. This makes finding the same landmarks much easier and reproducible. In addition, the option is available to have landmark read the GSF segmented spot list files and use this for finding spots when you indicate a position near a segmented spot. [4,8,10,12]

makjob
generates GELLAB-II Unix scripts. It requests a list of gel accession numbers for a subset of gels previously accessioned. It then asks a few questions regarding the type of experiment to be performed and generate Unix batch scripts to a) interactively landmark the set of gels, b) segment the gel images into GSF spot lists, c) pair GSF spot lists into GCF paired spot lists, d) merge the GCF files by constructing a PCG DB file and e) perform an initial statistical analysis of the PCG DB. A makjob run can be customized to perform some and not other specific analyses (see -info switch for makjob). See example of running makjob in the Examples. [12,13]

markgel
generates a Rmap image having specified a gel accession number and a SPSS-compatible (.sps) file generated by the cgelp2 program. The Rmap is the synthetic image generated by the projection of the set of spots specified by the SPSS file onto a copy of the gel image associated with the gel accession number. [4,9,12]

mosaic
generates one or more mosaic images having specified a particular Rspot number and a SPSS-compatible (.sps) file generated by the cgelp2 program. The imagee or plot of a Rspot for a set of gels is a composite image or graphic formed from panels from each gel arranged in a regular checkerboard pattern ordered by minimum spot density (protein concentration). The panels are taken from a subregion of each gel surrounding a particular Rspot. [4,9,12]

pgelrc
prints a user friendly form of the gel.rc GELLAB state file. This file contains the default names of various database files, directories, segmentation parameters, and information on the last data processed. It is used by all GELLAB-II programs upon startup If the file does not exist, then running pgelrc will run an interactive Question and Answer session to generate the initial gel.rc file and gellab directory tree. You can also find the status of the next free accession number and PPX file, that accession numbers are in the system, and what PPX files from the database are out on the disk. [12]

plotn
reads Universal Graphics Files (.ugf) produced by various GELLAB-II programs cgelp2, dwrmap, dendrogram. It is able to replot a .ugf on the same or different type of display as well as to plot the file on other devices (such as a Postscript laser printer). Used with tek2psG it can print plot files on a laser printer.

ppxcvt
convert different input picture file formats into the GELLAB Portable PiXture file format (.ppx). It can convert ASCII hex, decimal, or octal numbers as well as other binary formats with variable header, leading and trail number of pixels/lines, and variable number of lines. It also has input header optical density calibration conversion to PPX header format for selected input types (such as BioImage). Input images may be larger than the PPX output image in that case you have the option of sampling or averaging the data. Output images are normally 512x512 but may be specified to any size. Data may also be transformed by complementing, linear scaling, or performing a log transform to produce the 8-bit data currently required for the PPX file.

ppxodt
is a picture debugger for opening, reading pixels, 3x3 neighborhoods, and 18x18 windows of a .ppx file. Data may be viewed in hex, octal, or decimal. Individual pixels may be changed and the edited picture file saved.

ppx2ps
converts a .ppx image file to Postscript that can then be printed on a laser printer.

sg2gii
is a gel spot list segmentation program that generates a Gel Segmentation Files (GSF) from the image file associated with the gel accession number. The accession number and gel image are produced by the getacc gel image acquisition program. Image size defaults to 512x512 but any size image specified in the PPX file header (e.g., 1024x1024, 2048x2048, or 4096x4096) can be specified. [4,7,10,12,13]

tek2psG
converts Tektronix 4010/4015 input (such as is produced by plotn) to Postscript, that is suitable for printing on a PostScript laser printer. Tek2psG was derived from E. Moy's tek2ps program.

Xpix
is a general purpose X-windows X11R4 or later .ppx file interactive display program. It is controlled by the user moving and clicking a mouse to get menus selections and interact with the image(s). It can manipulate one or two images on the screen at a time with each image having its own real-time small zoom window (additional images can be kept in memory and be alternately displayed). It can perform general image processing types of operations where you can save transformed image disk files for later recall [27]. Xpix is symbolic-linked to either Xpix2 or Xpix11 that are later versions of the program.

4. Unix Directories and Support Files

Several Unix directories and support files are required for the operation of GELLAB-II. You need to add a path to the GELLAB-II executable and runtime documentation files that are kept in the system installation ~gelmgr/gellab/bin directory. Add the following line to your .cshrc Unix csh shell startup file:

set path = ($path ~gelmgr/gellab/bin)

Each user could have a subdirectory called ~/gellab for storing GELLAB-II project(s) data. Any other local user directory could alternatively be used. The master state file gel.rc contains the names of a number of other files as well as paths. These are indicated by a 'keyword=value' syntax in the gel.rc file. Both gellab and gel.rc can be put in your home directory. Alternatively, if you have several gel projects, create a separate directory for each one and then run pgelrc to create the required gel.rc, gel.id, lms.lm and gellab directory tree. These include the following entries that default to the following files:

gelFile= ~/gellab/id/gel.id the Accession Database file.
lmsFile= ~/gellab/lms/lms.lm the Landmark Set DB file.
spotListFile= ~/gellab/ann/spt.ann the Annotation DB file.

The default paths are:

ppnP1X= ~/gellab/ppx/ the original gel picture disk PATH.
ppnP2X= ~/gellab/aux/ the auxillary picture disk PATH.
ppnP3X= ~/gellab/tmp/ the temporary picture disk PATH.
ppnP4X= ~/gellab/pcg/ the PCG composite gel Database PATH.
ppnP5X= ~/gellab/gen/ the generated file Database PATH.

Note that .gsf and .gcf files as well as derived Rmap, mosaic, and segmented gel images are saved in the ppnP2X or aux directory. The composite gel database is kept in ppnP4X and generated files from an analysis of the composite database are in ppnP5X. To install a new GELLAB-II file system in a new account,

cd ~/
~gelmgr/gellab/demo/make-user.do

will create the required directory trees in your login account.

To access GELLAB-II, add

set path = ($path ~gelmgr/gellab/bin/`arch`)
to your .cshrc file. We assume you will run the C-shell csh or another compatible shell.

To create a new GELLAB-II sub-project 'prj' consisting of a separate set of gels:
cd (wherever you want to put the data)
mkdir prj
pgelrc

Then answer the questions (typing the RETURN key for the default is sufficient for most questions).

When using GELLAB-II with the X-Windows System, the GELLAB programs can be be invoked from interactive graphics menus. We currently use version X11R4 with the twm window manager with the .twmrc startup file. When a GELLAB program is selected, it may prompt you for a text response to supply the command line required. Currently several programs use X-Windows: Xpix, landmark, getacc, plotn and cgelp2.

5. Examples of GELLAB-II Commands to Run Programs

The following is a short set of examples illustrating some of the types of operations possible with GELLAB-II. No examples of cgelp2 are given because that is beyond the scope of this brief document and is covered in the Reference Manual [31]. See the detailed discussions in the papers listed in the References.

Example 1: print current state file if it exists. If not, it will ask you if the default values are acceptable and then create a gel.rc file and gellab directory tree for the project in the current directory.
pgelrc

Example 2: generate batch scripts given a list of gel accession numbers in file tst3.ccl, a project prefix ts3, four experimental classes and the rgel.
makjob -rgel:0324.1 -class:AML:ALL:CLL:HCL:HL-60 -accs:ts3.ccl -prj:ts3

Example 3: segment a gel into a GCF file and a 'z' segmented image spot file and then display it.
sg2gii 324.1
accppx 324.1 324.1 -P2:z

Example 4: given an accession number, display the gel in a Xpix window.
accppx 324.1

Example 5: given two accession numbers, display the corresponding gel image files in two Xpix windows.
accppx 324.1 369.1

Example 6: segment a gel as above but also generate the 'c' central core image and the original less the segmented spots image 'y' files. Then display the these two derived images.
sg2gii 324.1 -ctlcoreimage -restofimage
accppx 324.1 324.1 -P1:c -P2:y

Example 7: pair GSF spot list files for two gels into a GCF file.
cmpgl2 324.1 369.1

Example 8: pair GSF spot list files for two gels into a GCF file and also generate 'u' and 'v' labeled paired-spot image files. Then display the these two derived images.
cmpgl2 324.1 369.1 -MarkLabels
accppx 369.1 369.1 -P1:u -P2:v
or add the display switch to the same command:
cmpgl2 324.1 369.1 -MarkLabels -Xpix

Example 9: generate the two 'l' landmark set images for the two gels, but do not pair the gels.
cmpgl2 324.1 369.1 -onlyMarkLMSimages
accppx 324.1 369.1 -prefix:l
or add the display switch to the same command:
cmpgl2 324.1 369.1 -onlyMarkLMSimages -Xpix

Example 10: run cgelp2 on the command file batch script to create a PCG DB.
cgelp2 -f ts3cgl.gdo

Example 11: start the CGL database program on an existing database file.
cgelp2 -d ts3cgl.pcg

Example 12: generate an Rmap of gel 324.1 for SPSS file 'ts3s02.sps' procedure from the cgelp2 PCG DB program.
markgel 324.1 ts3s02.sps
accppx 324.1 -prefix:m

Example 13: same as above but display the Rmap in an Xpix window after it is generated.
markgel 324.1 ts3s02.sps -Xpix

Example 14: generate a mosaic image of Rspot 63 for SPSS file 'ts3s02.sps' produced from the cgelp2 PCG DB program.
mosaic 63 ts3s02.sps
accppx w00063.ppx

Example 15: same as above but display the mosaic in an Xpix window after it is generated.
mosaic 63 ts3s02.sps -Xpix

Example 16: Display a previously computed map and the mosaic image for a Rspot on the Rmap. Xpix window.
accppx 324.1 -P1:m w00064.ppx

6. References

1. Lemkin, P., Merril, C., Lipkin, L., Van Keuren, M., Oertel, W., Shapiro, B., Wade, M., Schultz, M., Smith, E. (1979) Software aids for the analysis of 2D gel electrophoresis images, Computers and Biomedical Research 12:517-544.

2. Lemkin, P., Lipkin, L. (1980) BMON2 - A distributed monitor system for biological image processing. Computer Programs in Biomedicine 11:21-42.

3. Lemkin, P., Lipkin, L., Merril, C., Shiffrin, S. (1979) Protein abnormalities in macrophages bearing asbestos. NIEHS Conf. Medical Aspects of Mineral Fibers. Environmental Health Perspectives 34:75-89, 1980.

4. Lipkin, L.E., Lemkin, P.F. (1980) Database techniques for multiple PAGE (2D gel) analysis. Clinical Chemistry 26:1403-1413.

5. Lester, E.P., Lemkin, P., Cooper, H.L., Lipkin, L.E. (1980) Computer-Assisted Analysis of Two-Dimensional Electrophoresis of Human Peripheral Blood Lymphocytes, Clinical Chemistry 26:1392-1402.

6. Lester, E.P., Lemkin, P., Lipkin, L.E., Cooper, H.L. (1981) Two-Dimensional Electrophoretic Analysis of Protein Synthesis in Resting and Growing Lymphocytes in Vitro, J. Immunology 126:1428-1434.

7. Lemkin, P., Lipkin, L. (1981) GELLAB: A computer system for 2D gel electrophoresis analysis. I. Segmentation and preliminaries. Computers in Biomedical Research 14:272-297.

8. Lemkin, P., Lipkin, L. (1981) GELLAB: A computer system for 2D gel electrophoresis analysis. II. Spot pairing, Computers in Biomedical Research 14:355-380.

9. Lemkin, P., Lipkin, L.(1981) GELLAB: A computer system for 2D gel electrophoresis analysis. III. Multiple gel analysis. Computers in Biomedical Research 14:407-446.

10. Lester, E.P., Lemkin, P.F., Lipkin, L.E. (1981) New Dimensions in Protein Analysis - 2D Gels Coming of Age Through Image Processing, Invited paper, Analytical Chemistry 53:390A-397A.

11. Lemkin, P.F., Lipkin, L.E. (1981) GELLAB: Multiple 2D Electrophoretic Gel Analysis, in Electrophoresis '81, R. Allen, Arnaud (eds), W. De Gruyter, New York. pp 401-411.

12. Lemkin, P.F., Lipkin, L.E.(1983) Database Techniques for 2D Electrophoretic Gel Analysis, in Computing in Biological Science, Elsevier/North-Holland, M. Geisow, A. Barrett (eds), pp 181-226.

13. Lemkin, P.F., Lipkin, L.E., Lester, E.P. (1982) Extensions to the GELLAB 2D Electrophoresis Gel Analysis System. Paper given at "Clinical Applications of 2D Electrophoresis", Mayo Clinic, Nov. 15-18, 1981. Clinical Chemistry 28:840-849.

14. Lester, E.P., Lemkin, P.F., Lipkin, L.E. (1982) A two-dimensional Gel Analysis of Autologous T and B lymphoblastoid Cell lines, Clinical Chemistry 28:828-839.

15. Lester, E.P., Lemkin, P.F., Lowery, J.F., Lipkin, L.E. (1982) Human leukemias: A preliminary 2D electrophoretic analysis, Electrophoresis 3:364-375.

16. Lester, E.P., Lemkin, P.F., Lipkin, L.E. (1983) States of differentiation in leukemias: A 2D gel analysis. In Chromosomes and Cancer: From Molecules to Man. Proceedings of 5th Annual Bristol Myers Symposium on Cancer Research. Academic Press, pp 226-245.

17. Lemkin, P.F., Lipkin, L.E. (1983) 2D Electophoresis gel database analysis: Aspects of data structures and search strategies in GELLAB, Electrophoresis 4:71-81. Presented at Argonne Workshop on Technical advances in 2D electrophoresis and clinical applications of the technique", Aug. 29-Sep.1, 1982.

18. Howard, R.J., Aley, S.B., Lemkin, P.F. (1983) High resolution comparison of Plasmodium Knowlesi clones of different variant antigen phenotypes by 2D gel electrophoresis and computer analysis. Electrophoresis 4:420-427.

19. Lester, E.P., Lemkin, P.F., Lipkin, L.E. (1984) Protein indexing in leukemias and lymphomas, NY Acad. Science 428:158-172.

20. Lemkin, P., Sonderegger, P., Lipkin, L. (1984) Identification of coordinate pairs of polypeptides: A techniques for screening of putative precursor product pairs in 2D gels. Clinical Chemistry 30:1965-1971.

21. Sonderegger, P., Lemkin, P., Lipkin, L., Nelson, P. (1985) Differential modulation of the expression of axonal proteins by non-neuronal cells and the peripherial and central nervous system, EMBO J. 4:1395-1401.

22. Lester, E.P., Lemkin, P.F. (1984) A 'GELLAB' computer assisted 2D gel analysis of states of differentiation in hematopoietic cells, In Neuhoff, V. (Ed.): In Electrophoresis '84, 1984. Basel, Switzerland, Springer-Verlag Chemie, pp 309-311.

23. Lemkin, P. (1985) PSAIL - A portable SAIL compiled translator for C environments, Computer Language 2:39-45. [Used in converting GELLAB-I to GELLAB-II]

24. Sonderegger, P., Lemkin, P.F., Lipkin, L.E., Nelson, P.G. (1986) Coordinate regulation of the expression of axonal proteins by the micro-environment, Developmental Biology 118:222-232.

25. Stoeckli, E.T., Lemkin, P.F., Kuhn, T.B., Ruegg, M.A., Heller, M., Sonderegger, P. (1989) Axonally Secreted Proteins: I. Identification of Proteins Secreted from Axons of Embryonic Dorsal Root Ganglia Neurons, EMBO J. 180:249-258.

26. Lemkin, P.F. (1988), PSAIL: A Portable SAIL to C Compiler - Description and Tutorial, SIGPLAN Notices Oct 23(10):149-171. [Used in converting GELLAB-I to GELLAB-II]

27. Lemkin, P.F. (1988) Xpix - An image processing system for X windows, Computers Biomedical Research 26:1-16.

28. Lemkin, P.F., Lester, E.P. (1989) Database and Search Techniques for 2D Gel Protein Data: A Comparison of Paradigms For Exploratory Data Analysis and Prospects for Biological Modeling, Electrophoresis, 10(2):122-140.

29. Lemkin, P.F. (1989) GELLAB-II, A workstation based 2D electrophoresis gel analysis system, in proceedings of Two-Dimensional Electrophoresis, T. Endler, S.Hanash (Eds), Vienna Austria, Nov 8-11, 1988, VCH Press, W.Germany. pp 53-57.

30. Lemkin, P. F. (1992) The GELLAB Papers - A Collection of Papers Describing the GELLAB-II System. NCI/FCRF, July 26, 1992.

31. Lemkin, P.F. (1993) The GELLAB-II 2D Gel Exploratory Analysis System. Reference manual, pp 677, August 1993.

32. Lemkin, P.F., Rogan, P., Automatic Detection of noisy spots in two-dimensional Southern Blots, Applied and Theoretical Electrophoresis 1991;2:141-149.

33. Amberger, A., Lemkin, P.F., Sonderegger, P., and Bauer, H.C. (1993): ECGF and heparin determine differentiation of cloned cerebral endothelial cells in vitro. Molecular and Chemical Neuropathology 20:33-43.

34. Myrick, J.E., Lemkin, P.F., Robinson, M.K., Upton, K.M. (1993): Comparison of the Bio Image VisageTM 2,000 and the GELLAB-II two-dimensional electrophoretic analysis systems. Applied & Theoretical Electrophoresis 3:335-346.

35. Wu, Y., Lemkin, P.F., Upton, K. (1993) A fast spot segmentation algorithm for 2D electrophoresis analysis. Electrophoresis 14:1350-1356.

36. Robinson, M.K.,, Myrick, J.E., Henderson, L.O., Coles, C.D., Powell, M.K., Orr, G.A., Lemkin, P.F. (1995) Two-dimensional protein electrophoresis and multiple hypothesis testing to detect potential serum protein biomarkers in children with fetal alchol syndrome. Electrophoresis 16(7):1176-1183.

37. Lemkin, P.F. (1995) Representations of protein patterns from 2D gel electrophoresis databases. In: Pickover, C., (Ed) The Visual Display of Biological Information. World Scientific Publishers, River Edge, New Jersey, pp 43-59.

7. Notes on Last GELLAB-II Release)

autopair was not released at that time.
Programs Xpix11 and Xpix2, getacc, landmark, plotn and cgelp2 work with X-Windows version 11.
Additional bugs were listed in the GELLAB-II Reference Manual [31].