DTASelect and Contrast Users` Manual
Contents
1. X1 Set highest 1 XCorr X2 Set highest 2 XCorr X3 Set highest 3 XCorr Table 2 Extended Spectrum Filters symbols indicate a numerical parameter follows while symbols indicate character or string parameters follow Extended filters only apply if they are specified accomplished with Stn M RK Stc MRK where the M RK requires that the residue preced ing this peptide in the database must be either the start of the protein sequence or a met arg or lys residue The Smn and Smx filters can be used to specify the minimum and maximum lengths of permissible sequences respectively The extended filters are not always useful but they can be helpful for seeking out particular sets of peptides 3 3 Program Flow DTASelect proceeds in three phases summarizing filtering and reporting Summarization is carried out by examining each out file in every subdirectory of the current directory The identifications are stored in memory Any sequence match with a DeltCN of 0 0 that is any identification that ties the best identification in XCorr will be stored for each out file Since multiple proteins may contain the same peptide some spectra will be associated with several proteins When all files have been read the out file information is sorted by locus This list is used to produce a list of genetic loci or proteins present with a list of evidentiary out files associated with each locus The summary of out files i2. Act or Phs labels respectively The colors used for labeling sequence ions are given in Table 8 If the fragment ions lose ammonia the resulting peaks are labeled with the same color as the source series with a to indicate that they are 17 Da away from the fragment ion Series Z 1 Z 2 a Green Light green b Red Light Purple y Navy Blue Light Blue Table 8 Sequence ion color coding in SpectrumApplet The series of letters at the top of the spectrum show how the peptide sequence aligns to the spectrum The colors of the letters show which ion series is being described see Table 8 If the peptide is triply charged four sequence rows will appear rather than the two appearing for other spectra If a sequence letter is black no ion in the spectrum corresponded to the peak to this amino acid residue s side The letters in the y ion row s are in reverse order relative to those in the b ion row s because the N terminal and C terminal ions appear in opposite orders by M Z SpectrumApplet presents the intensity of the base peak the tallest in the spectrum at the upper left of its display The program computes the proportion of the spectrum s intensity which is accounted for by fragment ions and reports it below the base peak intensity SpectrumApplet includes a control panel that does not appear in the DTASelect s GUI The options allow the labelling of a specific ion series rather than all simultaneously In additio
3. If the search has specified post translational modifications and these are included in the passed sequence SpectrumApplet can be notified of these symbols meanings by the following tags e DMMx The mass accounted for by a particular dynamic modification where x starts at 1 and increases by 1 for each additional static modification float e DMSx The symbol in the sequence corresponding to this this dynamic modification char In some cases peptide identification may be made with a mass modification for either the C or N termini of the protein or peptide These options can be passed to SpectrumApplet by the following options e CPepMod The mass modification on the C terminus of this peptide float e NPepMod The mass modification on the N terminus of this peptide float e CProtMod The mass modification on the C terminus of this protein sequence float e NProtMod The mass modification on the N terminus of this protein sequence float If the identification was achieved by use of monoisotopic masses rather than average masses the SpectrumApplet needs that information e AvgForFrag Should monoisotopic masses be used rather than average masses for calculat ing fragment ion positions boolean e AvgForParent Should monoisotopic masses be used rather than average masses for calcu lating the precursor mass boolean If the CGI needs to modify the display of the spectrm these three tags are available e LoMZ The minimum m z
4. L and u options are used to trim the list Finally the o option goes into effect A full description of these options follows Loci are retained if they pass either of two rules The first p keeps loci that have a sufficient number of distinct peptides remaining in evidence these peptides will all have different 8 sequences or charge states unless the t 0 option has been used The other rule r keeps loci which are supported by at least one peptide which has been identified redundantly enough to be deemed reliable For example a locus represented by two copies of the same spectrum will be retained under p 2 r 2 settings by passing the redundancy rule r but not the number of distinct spectra rule p The more typical way to pass this requirement is that a protein has three peptides when only two are required 7p 2 default setting The redundancy rule is set strictly enough by default 10 copies of a spectrum required that few proteins should pass it The M rule is quite similar to the p rule except that only peptides with dynamic modifications are counted and 0 are required by default These quota rules filter out large numbers of proteins for which scant evidence is present Two more filters remove proteins by simple criteria If the user has chosen the u option each locus will be examined to see whether any of the representative peptides are unique to that locus If at least one remains the locus is retai
5. Default INDIVIDUAL SPECTRUM FILTERS 1 1 8 Set lowest 1 XCorr 2 t 25 Set lowest 2 XCorr 3 3 5 Set lowest 3 XCorr d 0 08 Set lowest DeltCN c 1 Set lowest charge state C 3 Set highest charge state mz i Set minimum precursor m z MZ Set maximum precursor m z i 4 0 0 Set lowest proportion of fragment ions observed s t 1000 Set maximum Sp ranking m 0 Require peptides to be modified m 1 X Include peptides regardless of modification m 2 Exclude modified peptides y 0 X Include peptides regardless of tryptic status ye Include only half or fully tryptic peptides y 2 Include only fully tryptic peptides m x Keep N peptides discard all others v 0 X Ignore manual validation info v 1 Keep Y peptides discard N peptides y 2 Keep Y and M peptides discard N peptides ao Keep Y and M peptides discard N and U peptides Table 1 DTASelect Spectrum Filters symbols indicate a numerical parameter follows while symbols indicate character or string parameters follow Changes can be made at the command line or in a DTASelect params file Each spectrum must pass all filters The default criteria for DTASelect are designed to retain only the proteins for which ample evidence is present and to include results only from those out files that stand on their own merits The default criteria for DTASelect are conservative if a protein is listed under the de
6. are three fields that are used internally The first indicates where in this locus sequence this peptide matches 0 for the start of the sequence This value is used for determining sequence coverage One is a boolean value describing whether or not the peptide is a tryptic cleavage at both ends or is positioned at a terminus The other boolean value describes whether or not the peptide is unique to this locus These values are available for use by any program that reads the DTASelect txt file 4 Contrast Running DTASelect in a directory is the first step to analyzing its results against other directories Contrast draws on the information stored in the DTASelect txt files found in each of the specified directories to assemble a master protein list and to compare the different runs Contrast can be run in an arbitrary directory its configuration file specifies the directories to be included in its analysis 4 1 Execution Using Contrast requires a bit more preliminary setup than does DTASelect The user creates a file called Contrast params in a directory where results are to be stored This file specifies the directories to be included in the analysis and the sets of criteria to be used The Contrast script is then run The example Contrast params in Figure 9 compares the results of two samples under a common set of criteria that has been modified to increase the minimum necessary DeltCN One use of this technique would determine how the prot
7. aux Incorporate auxiliary protein information from AuxInfo txt XML Save XML report of filtered results DB Save in format for database import chroma Save chromatography report similar Save protein similarity table align Save sequence alignment report mods Save modification report help h Print this list of options here 4 Include only IDs in current directory Table 4 DTASelect Utilities evaluated by the above rules unless v 3 has been specified A quick summary of the command line rules appears in Table 5 Command Symbolic Longer Line Description Description v 1 N YMU Keep N peptides discard all others v 0 Ignore Ignore manual validation info Eu Y N Keep Y peptides discard N peptides v 2 YM N Keep Y and M peptides discard N peptides rd YM NU Keep Y and M peptides discard N and U peptides Table 5 Manual Validation Handling in DTASelect Secondly if multiple dta files are identified to match precisely the same sequence and charge state a spectrum redundancy exists If the user has chosen to do so only the best representative of these redundant spectra will be retained on the basis of highest XCorr or total intensity Either all t O orone t 1lor t 2 of each set of identical spectra will be kept Finally the loci in memory are purged Each protein must first pass either the r or p filter Next the e E 1
8. and 2 is a subset protein This option is likely to be most useful under p 1 single peptide per locus criteria or under very large sequence databases If the user does not indicate that parsimony should be used it is off by default The only proteins that will remain in the list are those that pass either r or p Next loci must pass the M rule followed by e E 1 L u and finally o if specified Again if manual validation criteria are put in place with the V option the locus filtering may be handled differently The summary of the peptide level v option in Table 5 is an accurate assessment of the locus level V option if one substitutes the word protein for peptide After filtering is complete the percentage of residues in each protein sequence that are rep resented by at least one peptide is calculated The output files are created and if the user has specified the GUI option the graphical user interface is displayed see Figure 2 If the user has specified the CGI option DTASelect will link to improved CGIs rather than those that ship with SEQUEST see Table 6 for more information Often the proteins in database have been categorized by biological function such as the MIPS annotation of the Saccharomyces protein database DTASelect can group proteins that are found in a sample by such classes when the c1ass option is employed The program will examine the contents of the Classifications txt file t
9. and the sequence database SEQUEST output out files are typically present in subdirectories off of the directory holding the sequest params file The database used for the SEQUEST search is identified from the sequest params file and it is expected to be in the same location as indicated in sequest params If a precompiled bin database was used the program will assume that a standard FASTA database with the same name except for the bin extension is present DTASelect has been tested with many versions of SEQUEST and will work with the out files produced by the SEQUEST in Bioworks 3 If a version of SEQUEST producing unified output files has been employed the directory may contain sqt files rather than out files DTASelect will check the current directory for sqt files and read these in preference to out files if they are detected DTASelect offers support for Matrix Science s Mascot algorithm output Identifications and configuration are read from MIME formatted dat files For information on this support see Ap pendix A Contrast requires that DTASelect has been run in each directory to be included in its analysis When DTASelect is first run in a directory it creates a file entitled DTASelect txt This file must be present in each directory included in a Contrast analysis Contrast reads from the Contrast params file in the current directory to determine the type of analysis to be performed The programs will expect the SEQUEST output
10. file names to follow this pattern DAT filename firstscannumber secondscannumber chargestate out Extraneous periods or different file name organization will likely confuse the program 3 DTASelect Once the dta files are extracted and SEQUEST has been run on the spectra DTASelect can be used The first run of DTASelect on any data set takes the most time During this first pass the program must read thousands of individual out files from the disk as well as sift through the entire selected sequence database Subsequent runs of DTASelect are far faster than its initial execution in a directory 3 1 Installation DTASelect installation begins when its files are copied to a directory on the computer where it will be run The DTASelect ini file see Figure 1 should be modified to configure which servers will be used for viewing spectra and sequence coverages which CGIs are available for example is Flicka installed rather than Web_retrieve and how long the locus names given by SEQUEST are different versions of the program crop the name to different numbers of characters A batch file or script to start the program should be created example batch files for Windows are included in the distribution On Windows machines this batch file should be called DTASelect bat and be stored in C WINNT or C WINDOWS On a Linux box the script could be stored in usr bin or another location located on the system path The script should look
11. least one peptide match present Peptide IDs provides the number of SEQUEST result files included in the analysis Individual out files may contain more than one sequence with a 0 0 DeltCN and identified se quences may match to more than one peptide Copies indicates how many identifications count of SEQUEST result files times number of top ranked sequences per file times number of proteins including each peptide listed are present before filtering The nonredundant count of proteins is the number of protein groups in the DTASelect out put while the redundant count is the number of individual proteins listed ignoring grouping The meaning of redundancy changes for the peptide and spectra counts for these counts individual sequences which were found in several different proteins are counted only once for nonredundant values but multiply for redundant counts If the t 1 or t 2 options have been used or the default setting used an individual spectrum may represent a group of several identical spectra As a result the number of spectra that passed the criteria will likely exceed the number of peptides reported Note that both peptide and spectrum counts are bogus when t O is specified the redun dancy of each peptide 1s still calculated correctly but the redundancy is not removed a necessity for these counts to be meaninful For both peptide and spectrum counts the nonredundant number counts each identification only once while the
12. of DTASelect s criteria across multiple data sets and criteria sets 1 Introduction As tandem mass spectrometry emerges as a primary tool for proteomics the complexity of the data produced through this technique has increased dramatically While SEQUEST aids enormously in the identification of individual spectra complementary tools to summarize SEQUEST results have lagged behind When faced with multidimensional chromatography results in particular the researcher needs assistance in grouping and selecting significantly matching spectra from more ambiguous results DTASelect and Contrast fill this niche DTASelect acts as an organizer and filter grouping together related spectra by protein and removing those that do not pass basic data quality criteria Contrast provides meta analytical capability helping the user to tune criteria to particular data dtabb u washington edu sets and to compare results from multiple samples Together these tools dramatically increase the scope of analyses possible through tandem mass spectrometry and SEQUEST This manual will assume that the reader is familiar with most features of SEQUEST operation A brief discussion of SEQUEST related files however is warranted SEQUEST is configured by modification of the sequest params file The information in this file includes the path of the FASTA sequence database and the differential modifications specified for the search SEQUEST s analysis of an individua
13. other Modified residues are indicated by being colored red or are shown in lowercase in the text report The starting and ending sequence positions of each contiguous region are included in the first cell of each row After each protein table four statistics are listed The Maximum Depth field shows the maximum number of peptides listed in any contiguous region for this protein not necessarily the largest number of peptides containing a particular residue of the protein The Peptide Residues Observed field sums together the length of each peptide observed for the protein and can be used to determine the mulitiplicity of sequence coverage for each protein The Sequence Coverage and Sequence Length show the same numbers as in the locus lines of DTASelect html Users who want to copy the peptides included in their DTASelect reports to another directory can do so with the copy option If SQT and MS2 files are found in the current directory DTASelect will create a new subdirectory called Subsets and create new SQT and MS2 files in it Only the identifications and spectra corresponding to the peptides shown in DTASelect html will be retained in the copied SQT and MS2 files If dta and out files have been used DTASelect will create a script entitled copylist bash on UNIX systems or copylist bat on Windows systems 3 5 DTASelect txt Databases The DTASelect txt file is created the first time DTASelect is run in a direct
14. redundant number counts each identification once for every protein group in which it occurs In some cases the ratio of the counts provide more information than they do individually For example the ratio of redundant to nonredundant proteins gives some indication of the extent to which the sequence database repeats itself If one wants to determine the number of spectra each unique reported peptide represents one can divide the nonredundant spectra count by the nonredundant peptide count Although understanding the source of each count can be difficult the information they provide can be useful Following the block of counts is a list of spectra which scored well in two different charge states For example a spectrum that contains y ions with minimal b ions may be matched to two overlapping peptide sequences when considered first as a doubly charged and then as a triply charged spectrum This list indicates matches that should be evaluated with a skeptical eye a spectrum should be only one charge state and scoring well at two states almost certainly means that one of the identifications is incorrect For larger sequence databases there will often be many proteins present which bear great similarity to each other DTASelect attempts to find pairs of proteins which are similar to each other for each sample and it reports this information in DTASelect pst txt when the similar option is used This file includes a row and column for each protein re
15. shown float e HiMZ The maximum m z shown float 25 lt applet code SpectrumApplet class CODEBASE http localhost DTASelect1 8 width 970 height 500 gt lt param NAME PreMPlusH VALUE 1159 72 gt lt param NAME PreZ VALUE 1 gt lt param NAME MZ1 VALUE 347 1 gt lt param NAME Inti VALUE 417922 0 gt lt param NAME MZ2 VALUE 348 3 gt lt param NAME Int2 VALUE 101197 0 lt param NAME MZ3 VALUE 349 7 param NAME Int3 VALUE 90863 0 gt Remaining peak parameters removed lt param NAME MatchSeq VALUE R IGSEVYHNLK S gt lt applet gt Figure 15 Basic HTML tags for SpectrumApplet e ShowPrecursorLosses Should losses from the precursor be shown float If other software is being used to recognize features of the spectrum peaks may be linked together by lines colored as the ion series are For example a sequence tag algorithm may predict a ladder of intensities and m z locations of several peaks The algorithm output may be displayed against the complete peptide identification e LadderLengthx The length of each ladder being described where x starts at 1 and incre ments by one for each ladder e LadderSeriesx The series of each ladder being described Series should be O for 1 b ions 1 for 1 y ions 2 for 2 b 3 for 2 y ions 4 for 1 a ions 5 for 2 a ions 6 for precursor related ions and 7 for display
16. DTASelect and Contrast Users Manual David Tabb December 16 2002 Abstract DTASelect provides a means by which complex SEQUEST results can be filtered orga nized and viewed A single sample may produce tens of thousands of tandem mass spectra Manually perusing and selecting SEQUEST matches among such a mass of data risks incon sistency DTASelect allows the user to set complex criteria for acceptance or rejection of individual spectrum results It also features rules for dealing with multiple identical peptide matches and for removing proteins that are insufficiently evidenced It provides its sorted and filtered summary as HTML and text documents for easy review and also offers several auxil iary reports DTASelect is a powerful tool for automatic analysis of complex mixture tandem mass spectrometry Contrast is a tool for differentiating multiple runs of data A user may run a control sample and experimental sample through the SEQUEST process and then use Contrast to highlight the differences in the protein compositions of the two samples An alternate use allows the user to determine the reliability of individual identifications in a sample by stepping through multiple criteria sets The tool can also be used with multiple data sets and multiple criteria sets simultaneously Results are stored in a unified web browseable summary and in HTML files representing each set under each set of criteria The Contrast program leverages the power
17. Dj V4 3 l4 N4 Dy 1 F 64 T j Ta Mis 4 41 T4 94 F4 J D4 N4 E Ys V4 D4 Aq E3 84 84 L4 Mb Mitts 1011748 0 My M 28 42 3 Lo m z Hi miz 214 19534 J Yl y CalcHz Diff 3 383 4 4 484 5 0 11 5 541 6 0 16 b8 6 688 7 0 44 3 8 b9 759 8 0 32 874 9 0 31 9 989 0 0 51 10 1102 2 0 67 44 11 1265 3 0 75 12 1364 5 0 12 13 1479 6 0 07 3 14 1550 6 0 15 15 1663 8 0 19 A b1 16 1750 9 0 69 8 b15 17 1838 0 0 36 18 1951 1 0 22 4 5 bb e bb bi 7 F bf2 4 i UR 2 hs M5 vag b18y1 4 vy P a 1 T pe i d AUT LORI UN LLL ea JL pe 400 500 600 00 800 900 1000 100 1200 1300 1400 1500 1600 1700 1800 1900 Applet started Figure 14 SpectrumApplet Sample The sequence shown is the doubly charged HELSSLAD VYINDAFGTAHR with a precursor M H of 2217 4026 with the path to the Mascot installation directory which should include a subdirectory entitled config in which Mascot dat can be found The first Mascot dat file supplies the database name DTASelect will expect that the individual spectra dta files will be stored in directories bearing the names of the corresponding dat files For example the spectra corresponding to F714280 dat should be found in a subdirectory entitled F714280 Mascot support in DTASelect was implemented relatively late in development As a result features associated with Mascot can be expected to change more rapidly than those relating to SEQUE
18. ST Additionally Mascot support may be less robust Your assistance in identifying areas where improvement is needed would be welcome B SpectrumApplet Interface DTASelect includes an applet which can be used to view spectra The SpectrumApplet displays spectra actively allowing zooming on particular M Z ranges and selective highlighting of fragment ion series The applet is accessible to any CGI which produces an appropriately formatted HTML file For an example of its use see Figures 2 and 14 The CGI produced HTML file referencing SpectrumApplet should specify its codebase as the DTASelect installation directory The spectral information is passed to the applet by the following required tags e PreMPlusH The precursor M H float e PreZ The precursor charge state integer 24 e MatchSeq The sequence corresponding to this spectrum String e MZx The M Z value for this peak where x starts at 1 and increments by one for each peak float e Intx The intensity value for this peak where x starts at 1 and increments by one for each peak float Many HTML tags are available as options in SpectrumApplet calls If nonstandard masses have been used for amino acids these options pass the new mass to the applet e SMMx The mass to substitute for a particular amino acid residue where x starts at 1 and increases by 1 for each additional static modification float e SMRx The residue which has the substituted mass above char
19. ascot results DTASelect must translate from these values to SEQUEST equivalents The following fields are affected by these differences 22 XCorr Mascot s Ions Scores are reported in the place of SEQUEST s XCorr Ions scores have a much broader range than XCorr ranging upwards of 60 and down to around zero DeltCN Mascot s Ions Scores should be evaluated in relation to the number of peptides present in the sequence database near the observed precursor mass Qmatch score DTAS elect looks up the Qmatch and Ions Scores and reports DeltCN as 107 10 Q where D is DeltCN I is Ions Score and Q is Qmatch The resulting number indicates the chance that each peptide is a random match A DeltCN of 1 0 corresponds to a 1 20 chance of random match A DeltCN of 5 0 implies a 1 100 1 20 divided by 5 0 chance of random match M H Mascot reports intact peptide masses while SEQUEST reports peptides masses with a single added proton M H masses The M H masses of peptides are calculated from Mascot results to match their SEQUEST equivalents Rank by Sp Mascot does not use preliminary scoring to create a candidate list As a result DTASelect reports all Mascot identified peptides with a preliminary score of 1 Filename Currently DTASelect will only detect filenames correctly if Mascot has been used on dta files Total Intensity DTASelect records a zero for the total intensity of each Mascot identifica tion This value coul
20. d and is followed by the number of spectra listed the number of spec trum copies present for this protein the sequence coverage specifically what percentage of the residues are represented in observed peptides length of the database sequence in residues molec ular weight approximate pl and descriptive name The peptide identifications are shown below the protein to which they belong Each out file s information includes whether or not the peptide sequence is unique to this locus indicated by an asterisk at the start of the line the file name the XCorr and DeltCN scores the observed M H mass the rank by preliminary scoring the percentage of ions expected that were found the number of times this peptide was identified and the sequence of this peptide Peptides that are found in mulitple loci have a series of plus symbols after the sequence Each symbol links to a different locus within the HTML file that contains that peptide Peptides that have been identified by the SEQUEST PHOS algorithm rather than normal SEQUEST are indicated by the letter P at the start of the line The peptides are listed in the order in which they align to the full protein sequence with ties broken through ordering by observed M H mass Following the list of peptides some protein groups will show a list of similarities linking to other protein groups containing similar peptides The first number listed after each locus is the number of peptides in this l
21. d be calculated from the intensities for each query in the dat file but DTASelect does not yet implement this Ion Proportion DTASelect reports the following for this measure M L 1 X where M is the reported Ions Matched score L is the length of the peptide sequence and X is 2 for singly charged and doubly charged peptides or 4 for triply charged peptides Rather than use the default settings for DTASelect Mascot users should use a command line like Mascot d 1 0 1 0 0 2 0 0 3 0 0 This should filter out peptides with worse than 1 20 chances of random matching regardless of charge state Mascot files are organized into folders based on the date when they are created To select a group of Mascot dat files which should be grouped for an individual DTASelect make use of the DTASelect params file Create the file in the directory where the DTASelect is to be run first line of DTASelect params must give the default parameters to be used for DTASelect in directory see the above paragraph for recommended defaults The next line should read 1 List Subsequent lines should contain only the path and filename for each dat to be inclu Comment lines should start with characters The this DAT ded DTASelect must be able to find the Mascot dat file to determine the path of the sequence database used for searching The DTASelect ini file should contain a line labeled mascot p 23 at 29 La 84 384 L3 4
22. d then by locus name Each row normally represents one protein however if multiple proteins have identical se quence coverage they are grouped together as in DTASelect s output The counts at the end of each table count such groups as a single protein The values representing sequence coverage in individual data sets will be those for the first protein in the group colored red Each protein s cumulative sequence coverage however is shown in the total column At the end of each subtable is a count of the loci in that group Near the end of the Contrast html file is a short table providing counts for each combination of presence and absence and a total of loci presented The example results in Figure 13 correspond to the example Contrast params in Figure 10 and show the summary table for two samples analyzed under two different conditions The last item in Contrast s report is a two way comparison table Each row and column repre sents a data set The protein lists for each pair of columns in the analysis are compared and the percentage of proteins found in the row s data set which are also found in the column s data set is shown in each cell If two samples A and B are being compared under the same criteria for example there will be two rows and two columns of percentages in the table The cells for row A column A and row B column B each show 100 any data set s list of proteins is identical to itself The cell for row A column B will
23. e Options section and the proteins present in any combination of the Control sample with any criteria set will be hidden Only protein counts are displayed for hidden groups If multiple samples are all named Control they can be removed by a single hide statement If multiple samples or criteria sets are to be hidden multiple hide options one on each line should be employed Master specifies a particular data set to be used as the definitive protein list for the Contrast run Master should be followed by a directory alias and then a criteria set alias such as Master Yeast2 Strict from Figure 10 This specifies that only the proteins found in Yeast2 Strict will be included from Yeast Loose Yeast Strict and Yeast2 Loose This option would be of use if one were investigating lower scoring peptides in evidence for a set of confidently identified proteins 19 The merge option will combine together all the listed DTASelect txt files into a unified DTASelect txt created in the directory where Contrast is run DTASelect or Contrast can subse quently be run against this newly created file although the links to spectra and SEQUEST output files will be made on the assumption that the relative paths to spectra are unchanged from the original DTASelect txt files Contrast s verbose mode details the presence and absence of individual peptides for each protein listed across all data sets showing the high
24. ed ions to appear in black e LadInty The intensity of a peak in a ladder y should begin at 1 and increase by one for each point If two ladders of three peaks are shown the first ladder is points 1 3 and the second ladder is points 4 6 e LadMZy The m z value for a peak in a ladder The parameters can be passed in any order The sequences should be passed as they appear in SEQUEST results R IJIALSRPNAYM FK Y with flanking sequence characters is required rather than IALSRPNAYM FK Figure 15 shows an example of a valid call to SpectrumApplet In this example The precursor s M H is 1159 72 and it was observed singly charged One static modification was present with cysteines being expected at 160 139 instead of 103 daltons Any occurence of in the sequence corresponds to a residue modification of 16 daltons an oxidation Since neither a C nor a appears in the sequence these parameters could have been omitted 26 The sequence fragment ions highlighted in the spectrum are the most intense falling within 0 75 M Z of a calculated fragment ion position the DaughterDB software manual gives a more detailed description of how the M Z positions of fragment ions are calculated The precursor ion if any remains is labeled with a brown tick mark below the x axis and if ions appear at the expected positions for ammonia water acetyl or phosphate neutral loss from the precursor they are colored brown and marked with
25. ein complement of an experimental sample differs from a 17 control Included Directories Yeast data dtabb Yeast Yeast2 data dtabb Yeast2 Criteria Sets NewDCN d 0 1 Options Figure 9 Sample Contrast params for Differential Analysis The first directory is given the alias Yeast and the second is aliased as Yeast2 The criteria set is nicknamed NewDCN These abbreviated titles are used to name each combination of sam ple and criteria hereafter described as a data set In this case two data sets would be created namely Yeast NewDCN and Yeast2 NewDCN Note that the criteria modifications are listed just the same as they would appear in DTASelect s command line The Contrast params found in Figure 10 is a more complex demonstration Two samples will be compared as before but this time two different sets of criteria are used Each combination of sample and criteria is evaluated separately so four different columns will be appear in the final output Included Directories Yeast data dtabb Yeast Yeast2 data dtabb Yeast2 Criteria Sets Loose p 1 d 0 0 Strict Options Figure 10 Sample Contrast params for Complex Analysis Strict has no modifications listed and will be set to the default criteria In this case the Yeast Loose data set is likely to retain more proteins than Yeast Strict because of the eased restriction on peptides pe
26. eptides 0 false Remove proteins that are subsets of others mw Set minimum protein molecular weight MW Set maximum protein molecular weight pi Set minimum protein isoelectric point PI Set maximum protein isoelectric point exer Remove proteins with IDs matching this string E Include only proteins with IDs matching this string e S Remove proteins with descriptions including this word eI Include only proteins with descriptions including this word M 0 Set minimum modified peptides per locus criterion r 10 Show all loci with peptides that appear this many times p 2 Set minimum peptides per locus criterion Table 3 DTASelect Locus Filters symbols indicate a numerical parameter follows while symbols indicate character or string parameters follow Filter changes can be made at the command line or in a DTASelect params file Each locus must pass either r or p and then any e E 1 L u and o options specified Option UTILITIES nofilter n Do not apply any criteria COpy Create script to copy selected spectra and IDs or subset SQT and MS2 files GUI Report through GUI instead of output files compress Create IDX and SPM files from spectra CGI Use the replacement CGIs instead of the classic ones Mascot Draw peptide IDs from Mascot dat files instead BE Produce Bird s Eye view of proteins found Class Classify proteins according to Classifications txt
27. escription Class Incorporate protein classification information in Contrast html Database Create a new FASTA database containing the selected proteins Hide Do not show proteins found in a particular sample Master Include only proteins found in this data set Merge Create a DTASelect txt file integrating these samples Verbose Show patterns of presence and absence for peptides as well as proteins Table 7 Valid Options for Contrast params The class option is similar to DTASelect s class option Classifications txt is read and the master protein list assembled in the current Contrast is classified according to the file s contents The proteins will still be grouped according to the data sets in which they are found but within those groups they will be sorted by the assigned classifications The database option creates a new sequence database Contrast fasta containing only the proteins listed in the Contrast output The program will search for the listed proteins in each of the databases corresponding to the samples included in the comparison Such a reduced database can dramatically reduce the time required for subsequent SEQUEST searches of the spectra with additional post translational modifications The hide option helps reduce the size of Contrast output If only the proteins present in the Experiment sample but absent in the Control sample are of interest include hide Control in th
28. est XCorr for each peptide in each data set Verbose mode output can be substantial in length with even moderately complex samples 4 2 Output Interpretation Contrast s output is a collection of files The primary ones are Contrast html and Contrast txt These files consist of a header showing the included directories as well as the criteria sets for the comparison followed by a series of tables one for each combination of presence and absence across the data sets Each table has a column for the locus name one for each data set one for cumulative sequence coverage and one for the descriptive name for each locus The other reports are the html and text files from DTASelect for each combination of directory and criteria set For each row in a Contrast table the cell for each data set may either be blank if the locus is not found there or may hold a percentage indicating the sequence coverage from peptides passing that set of criteria in that sample The percentages are hypertextual links to the correct locus in an HTML file produced by DTASelect for that data set The total column holds cumulative sequence coverage for all peptides found in all data sets for this protein and links to a view of the protein s sequence coverage The partial results in Figure 12 correspond to the Contrast params in Figure 9 and show a group of proteins present in Yeast but absent in Yeast2 Proteins within a single table are sorted by classification if used an
29. faults there is strong evidence that it is present in the sample The default criteria are listed in the second column of Tables 1 and 3 Many of the filters bear further investigation The individual spectrum filters are fairly straight forward setting minimum or maximum limits on the basic parameters of each identification The extended filters however may not seem terribly useful at first glance If one is attempting to find spectra that SEQUEST does not identify well the maximum XCorr settings can come in handy The sequence based filters are helpful for special uses For example if one wants to exclude peptides that result from incomplete tryptic digestions using Sec RK will manage this task Identifying peptides that carry certain motifs or which constitute affinity tags can be accomplished with the Sip option Creating customized cleavage filters can be handled with Stn and Stc Filtering to include only peptides that would result from a combined CNBr Trypsin digest can be Option EXTENDED FILTERS off by default Sic Sequences must contain all of these characters excludes C terminal residue Sip Sequences must contain this pattern Sec Sequences must not contain any of these characters excludes C terminal residue Stn Preceding residue must be one of these Stc C terminal residue must be one of these Smn 4 Sequence must be at least this length Smx 4 Sequence must be no longer than this
30. from the compressed files or from dta or ms2 files Clicking on the name of a spectrum in the list will display it in the viewing window For a description of the spectrum view see Appendix B Name Description Author Show Displays spectra via SpectrumApplet Hayes McDonald SeqCov Displays depth of sequence coverage Johannes Graumann Evalocus Allows changes in locus validation states Johannes Graumann Evalpeptide Allows validation of peptide identifications Hayes McDonald Table 6 Auxiliary CGIs designed for use with DTASelect Note that these programs are not included in the base distribution of DTASelect and may require separate licenses for use For more information about the SpectrumApplet spectrum viewer see Appendix B 10 example If class is used with the DB option a file entitled DB Classes txt is created along with the others class 1 Huge Proteins class 2 Big Proteins class 3 Sizeable Proteins class 4 Small Proteins class 5 Tiny Proteins YLR106C 1 YKR054C 1 Figure 3 Sample Classifications txt file Six classes are described numbered 1 6 up to 126 are possible Individual proteins are assigned to a group by having their locus names followed by the number corresponding to the appropriate group Here YLR106C and YKR054C are assigned to group 1 Huge Proteins More loci would follow in the full file Proteins which are not associated with a group are assigned to group 127 the unclas
31. l represent the second Both columns are compared using the same set of criteria so any particular sample appears in just one column If researchers were trying to determine which proteins appeared in sample one but not in sample two they could follow a link to the table grouping together those proteins showing a percentage in the first column but nothing in the second Conversely if they wanted to determine how many proteins were in the second sample but not in the first they could jump to that group showing a blank in the first column but a percentage in the second Proteins that appeared in both samples would be sorted together in the group at the top of the page If they were only interested in a count of how many proteins fell in the first but not the second they could follow the link to the bottom and read the numbers from the summary table rather than hunt through the rest of the document It is more complex to interpret a multi sample multi criteria analysis In Figure 10 two differ ent samples and two sets of criteria are present As a result four columns appear in Contrast html in addition to the three columns showing counts and percentages for each group One can quickly learn that 93 different proteins appeared in both samples under both sets of conditions the count rises to 117 if one counts indistinguishable proteins separately The 93 proteins comprise 8 6 of the total 1079 proteins To learn how many loci appeared in the Yeast2
32. l spectrum dta file is recorded in file with the extension out These files include loci and sequences which match to this spectrum as well as the scores characterizing each match such as XCorr deltCN and Sp A single MudPIT LC LC MS MS experiment may yield thousands of out files in several subdirectories DTASelect groups together these SEQUEST results to produce a unified report Comparing the protein content of two samples has been a daunting task Because DTASe lect automates the process of filtering the protein lists however the challenge is greatly reduced Contrast uses DTASelect s filtering capacity to create protein lists for multiple samples and then compares the proteins found in each What could otherwise be a nightmarish session with a large spreadsheet becomes as simple as specifying the directories to be compared and waiting a minute 2 Requirements DTASelect and Contrast can be run on any Java 1 1 compliant virtual machine VM These VMs are widely available for many platforms The most widespread VM installs with Microsoft s In ternet Explorer 4 0 and later If you use Microsoft Windows and want to determine whether or not you have this VM installed type jview at a command line The IBM Java 2 Developer s Kit for Linux and Windows v1 3 was used for compilation and execution during development DTASelect draws information from three sources while running SEQUEST s output files the sequest params file
33. llow the appropriate option If an option is left off the command line its default value is used Users may specify alternate default options in a file entitled DTASelect params in the directory where the dta files are located This file s format is a single line of options entered as they would be on the command DTASelect configuration file List the SEQUEST CGI server s name or IP address below server name localhost Which CGI will display spectra spectrum display cgi shl displayions exe Which CGI will display SEQUEST output files output display cgi shl showout pl Which CGI will show sequence coverage sequence coverage cgi shl flicka pl Which CGI will handle protein validation protein validation cgi bin EvalocusA Which server and CGI should handle BLAST searching BLAST handler www ncbi nlm nih gov 80 blast Blast cgi What additional options should be sent to BLAST BLAST args Is this a Windows or UNIX computer server type Windows When newer CGIs are used which server should be used cgi server name localhost What is the length at which my SEQUEST truncates locus names locus length cutoff 21 Where is Mascot installed mascot path c inetpub mascot Figure 1 Sample DTASelect ini file line It is possible to run DTASelect without any specified options all of the default values are used in this case Valid options are listed in Tables 1 2 3 and 4 Option
34. mitted and can be opened in most spreadsheets and databases To compare the proteins found in a run against the full set of proteins in a database one can use the MoreFASTA tool The program takes as an argument the name of a FASTA database If given the database yeast orfs it produces a file called yeast orfs DB Each line of the DB file is similar but not identical to the format of the DB Proteins file and contains the locus name sequence length molecular weight calculated pI description and residue composition of each protein in the database 11 Whole Proteome View l Whole Proteome View Huge Proteins EE Big Proteins Piddling Proteins NEEZEEEEEE EN Small Proteins EEEEEHENENE HEEEEEEEEH EEEEEHEEHEH HHEHEHEHHH EH an NH HEEBHEHHEEH EHEUBEBEEEEI MEHEEEEEEE H Tiny Proteins Unclassified EENHEEEEEENH HEO CHEREE BER NEEE Total Proteins 2156 Sequence Coverage Legend m100 0 87 5 075 850 07 Figure 4 Sample BirdsEye Graphic for Yeast Proteome Sample This example uses classifications to separate the proteins of the yeast orfs database into six classes on the basis of mass leaving a few of the smallest proteins unclassified 12 Set title for Bird s Eye Viewer below Title Whole Proteome View Should the title appear Show Title true Set size of dot for each protein Dot Size 8 Should we show a legend Legend true Should we show an overall count Overall Stats true Sh
35. n users can specify high and low M Z cutoffs to view only the region of the spectrum after 500 M Z type 500 into the Lo m z box and hit enter The y axis does not change when the x axis is zoomed Below the control panel users can check the correspondence between observed and predicted fragment ion M Zs for a particular ion series Ions that do not match to the spectrum are in black The middle column of numbers shows the M Z at which each fragment ion was expected to appear and the column at the right shows the discrepancy between the expected and observed positions The default list shown is for 1 y fragment ions If series are switched off by the checkboxes the program will instead show 1 b 2 y or 2 b ions 24
36. ned Otherwise the locus is removed The exclude option e may be used to remove groups of proteins with names containing the specified key phrase while the include name option E requires that all proteins match a particular string These options are most useful when the sequence database has been created by merging together two disparate ones and the database IDs from each component database use different patterns The exclude option could be used for example to remove all proteins with IDs that start with gi by specifying e gi on the command line or in Contrast params The 1 and L options are similar except that they are applied to the locus descriptions rather than the locus names themselves and are case insensitive Not all databases feature extensive descriptions but someone searching for proteins with kinase in the description could use these options to list just these proteins All of these filters are only in effect if explicitly specified The final locus filter is more complex to describe Proteins are then grouped by combined sequence coverage If two different proteins have identical sequence coverage the proteins will be grouped together for DTASelect s output If the user has selected the parsimony o for William of Occam option subset proteins are removed A subset protein is one for which all peptides are found in another protein For example if protein A has peptides 1 2 and 3 protein B with only peptides 1
37. o determine how to group proteins see Figure 3 for a brief 9 f DTASelect G4 E4 L4 A F4 Nq L3 64 V4 V4 Gy L3 N4 Py 2342127 0 52 9 200 300 Locus DC sc SC L M pl Descriptive Name Filename XCorr DeltCN M H SpR Ion Sequence RFP YORI84U 4 4 15 2 395 43416 6 5 SER1 Chr XV from 679356 680543 051801ym deca2 07 2717 2717 3 3 577 0 2923 3061 9 1 23 1 1 M PTPVYLQQAAK DLINFNDIGLGIGEISHR S 051801ym deca2 06 1540 1540 2 3 22 0 2859 1984 5 7 38 2 1 K DLINFNDIGLGIGEISHR 5 051801ym deca2 05 0859 0859 2 3 669 0 52 1578 7 1 71 4 1 K IAPAGYLVTGSWSQK S 051801ym deca2 07 2045 2045 2 2 948 0 395 1938 0 1 46 9 1 K AVQNLVDFIKEFAEEKNA ORFP YHLO33C 3 4 14 8 256 28125 10 0 RPL8A Chr VIII from 35253 36023 reverse complement ORFP YLLO45C 3 4 14 8 256 28112 10 0 RPLS8B Chr XII from 47858 48628 reverse complement 051801ym deca2 06 0819 0819 2 4 995 0 388 2016 5 2 55 3 1 K TSAVAALTEVRAEDEAALAK L O5180lym deca2 06 0732 0732 2 4 037 0 2782 1929 8 1 53 1 2 K LVSTIDANFADKYDEVK K O5180lym deca2 06 0685 0685 2 4 202 0 3878 2057 9 1 52 9 1 K LVSTIDANFADKYDEVKK H ORFP YGR282C 8 ll 14 75 313 34119 4 5 BGL2 Chr VII from 1057780 1058721 reverse complement 051801ym deca2 05 0947 0947 1 1 863 0 1962 1160 7 1 60 0 2 A IGELAFNLGVK N 051801vm deca2 10 1131 1131 1940 7 1 73 5 2 K IKESTVAGFLVGSEALYR N Figure 2 Sample of DTASelect s GUI displaying a spectrum Spectra can be viewed directly
38. ocus matching the other The second number is the count of peptides in this locus not matching the other The similarity links appear only when the similarity score as described below is zero or greater U YGLI47C 2 5 7 9 191 21604 9 7 RPL9A UYNLO67W 2 5 7 9 191 21692 9 7 RPL9B yeast 2052 2052 2 4 0341 0 3996 1807 64 167 9 2 R YVYAHFPINVNIVEK D yeast 1646 1646 2 2 5635 0 272 1381 86 172 7 3 Y AHFPINVNIVEK D Similarities YLLO24C 1 1 U YGRO34W 2 2 15 5 129 14639 10 3 RPL26B yeast 0695 0695 2 2 865 0 2723 1257 73 175 0 1 R KAYFTAPSSER R yeast 0342 0342 2 3 5671 0 2422 1101 44 193 8 R RDDEVLVVR G Figure 7 Sample DTASelect html Fragment Locus lines supply the current assigned manual validation status U means no status assigned the locus name the number of peptides listed below the total number of spectrum copies representing this locus the percentage of sequence coverage the length of the full protein sequence the average molecular weight of the protein the pland the descriptive name of the locus from the database 14 The end of the file is a table of several counts The unfiltered row describes the data set before filtering comes into play and the redundant and nonredundant rows describe the data appearing in the DTASelect report The Unfiltered row s three counts are calculated directly from the DTASelect txt file with no filtering used at all Proteins gives the count of proteins that have at
39. ory This file is useful beyond its capacity to speed up subsequent runs of the program In essence the file embodies the primary features of every out file in the subdirectories In addition the file includes the descriptive 16 gi 117385 sp P02511 CRAB_HUMAN ALPHA CRYSTALLIN B CHAIN 19 RRPFFPFHSPS 2 8212 07160134Sela 18 1284 1284 2 1 1 80 0 RPFFPFHSPSR 3 2679 07160134Sela 18 1217 1217 3 1 1 HSPSRLFDQFF 4 738 06130134Ssub 14 1615 1615 2 1 4 HSPSRLFDOFFGE 5 301 06130134Ssub 14 1462 1462 2 1 47 21 HSPSRLFDQFF 2 7263 06130134Ssub 16 1923 1923 2 1 2 80 0 HSPSRLFDOFFGE 4 4624 06130134Ssub 16 1853 1853 2 1 4 35 RLFDOFFGEHLLESDLFP 2 599 07160134Sela 14 1842 1842 2 1 1 80 0 Figure 8 Sample DTASelect mods html Fragment name of each locus identified from the database Perhaps its most useful feature though is its organization Each locus is listed with the out files that substantiate it This information is recorded before criteria are used against the out results the DTASelect txt file is comprehensive The file is tab delimited text Lines describing loci begin with L while those starting with D provide the information associated with a particular spectrum s match results Most fields in each D line are printed for the HTML file and do not require additional explanation At the end of each of these lines however
40. ould we show a count for each class Class Stats false What font size should the title be Title Font Size 18 What font size should the other text be Other Font Size 14 Should we use a white background rather than black Reverse Colors true Figure 5 Sample BirdsEye ini file Modifications to these values will be reflected the next time DTASelect is run with the BE option int Rank float Score YHR174W 3 4 5623 YLRISAW 2 7 5342 Figure 6 Sample AuxInfo txt file Two extra fields are added for each protein in this list The first is an integer and the heading used for the column will be Rank The second is a floating point number and its heading will be Score Two proteins are listed each with its associated values in the order they re listed in the header of the file Additional proteins could added on subsequent lines 13 3 4 Output Interpretation The resulting list of proteins and supporting peptides is stored in DTASelect html and DTASelect filter txt The file headers include the version of DTASelect used the directory in which it was run and a table showing the criteria in place for this analysis An example of the DTASelect html file s body can be found in Figure 7 Multiple loci are grouped together if each is identified by the same set of peptides Loci are shown in order of de creasing sequence coverage and are grouped by classification if class has been used Each locus name is printed in re
41. r locus and DeltCN Despite the fact that these data sets come from the same sample Contrast evaluates them separately The Contrast algorithm is limited to 63 columns the number of samples multiplied by the number of criteria sets cannot exceed 63 The availability of RAM and the width of the browser windows may be more substantial limits for the program In some cases such as the comparison of different peptide identification algorithms it may be necessary to apply one set of criteria to one set of results while applying a different one to another set This can be achieved in a section of Contrast params titled Explicit Mappings This section should always follow Criteria Sets and preced the Options section if one is used Each line in this section should start with the alias for an included directory Aliases for criteria sets to be applied to each directory should follow separated by spaces or tabs In the above example 18 one could specify that the Loose criteria should be applied to the Yeast sample while the Strict criteria should be applied to the Yeast2 sample by including the contents of Figure 11 in Contrast params Explicit Mappings Yeast Loose Yeast2 Strict Figure 11 Sample of explicit mapping between criteria sets and directories The final section of Contrast params Options can contain lines for any of several directives see Table 7 These keywords are case insensitive Option D
42. roup all the peptides from each DAT into sets of 100 scans a peptide numbered 082500dtabbB05 1625 1625 2 out would be in the seventeenth group 1625 100 16 25 which rounds up to 17 A graph of the numbers after the fifth column for each row then will be something like a total ion current taking into account only the peptides that remain after DTASelect filtering The information in DTASelect chroma txt is helpful in determining the effectiveness of the chromatography employed for a run Research focusing on post translation modifications can benefit from DTASelect s modification report This report generated when the mods option is used yields information about proteins with at least one modified peptide present see Figure 8 Available in web DTASelect mods html and text DTASelect mods txt formats the report enumerates the residues in each protein for which at least one dynamic modification was observed When multiple peptides are observed for a particular modification a sequence alignment of the identified sequences is produced with a marker to show the modification s position The alignment report can be helpful in showing which regions of a protein s sequence are most amenable to mass spectrometry The report generated when the align option is in effect creates a table for each protein Each row in the tables represents a region of contiguous sequence coverage and shows how the peptides in these regions align against each
43. s written to the DTASelect txt file Subsequent runs of DTASelect will not read the out files but rather will import the DTASelect txt file The next stage of DTASelect removes low scoring and redundant information The three step process removes spectra that do not meet the basic criteria ejects redundant spectrum copies and then removes loci that do not pass the necessary criteria First the individual spectrum filters are applied to each peptide If a peptide has an insufficient XCorr score for example it will be removed The majority of criteria available in DTASelect are parameters describing the removal of individual spectrum results at this stage A spectrum must pass all of these filters to be included If peptide level validation has been selected via the v option individual spectrum filters may be bypassed for some or all peptides If a peptide has not been validated then its spectrum is Option Default LOCUS FILTERS f 0 Do not purge duplicate spectra for each sequence i Purge duplicate spectra on basis of total intensity wb 2 X Purge duplicate spectra on basis of XCorr EXE Keep N peptides discard all others V 0 X Ignore manual validation info e Keep Y proteins discard N proteins c Keep Y and M proteins discard N proteins V 3 Keep Y and M proteins discard N and U proteins u false Include only loci with uniquely matching p
44. sample but not in Yeast however would require adding together the results where Yeast2 was positive in both columns and Yeast was negative in both columns When one increases the complexity of Contrast s output it becomes more difficult to answer simple questions from its results Keeping the Contrast params file as simple as possible will pay off in simplicity of understand ing the result A complex analysis can often be split into several simple analyses and it may be beneficial to do so 5 Summary DTASelect and Contrast are powerful tools for the sifting meaningful results from complex SE QUEST results By providing a framework for the uniform automated application of identifica tion acceptance criteria DTASelect greatly diminishes the work required to identify worthwhile sequence identifications from the large numbers of spectra typically produced by liquid chromatog raphy tandem mass spectrometry Contrast provides meta analytical tools for the comparison of multiple samples to each other and for the examination of samples under multiple sets of crite ria Just as SEQUEST improved the throughput of peptide spectrum identification DTASelect and Contrast improve the throughput of SEQUEST data analysis A Mascot Support in DTASelect DTASelect was originally created to analyze SEQUEST output As a result the pieces of informa tion stored for each peptide and filters designed to screen them are characteristic of SEQUEST To read M
45. show the percentage of proteins present in A that are also found in B The cell for row B column A will show the percentage of proteins present in B that are also found in A If one sample has many more proteins than another it may be the case that it represents a large percentage of other samples proteins Each cell in this table is color coded high percentages map to red or orange and low percentages map to yellow or green 20 Locus Yeast Yeast2 Total Description NewDCN NewDCN YBRIS81IC 11 11 RPS6B YBLO90C 11 RPS6A YCLO37C 11 11 SRO9 YCLO043C 17 17 PDII YDLO0SSC 8 8 PSAI YDLOSIC 28 28 RPPIA YDR099W 15 15 BMH2 YER177W 15 RPL23B YDR382W 49 49 RPP2B YDR450W 16 16 RPS1I8A YMLO026C 16 RPS18B YELO26W 31 31 28 X X Figure 12 Sample Contrast html for Differential Analysis R Count NR Count Percent Yeast Yeast2 Yeastl Yeast2 Loose Loose Strict Strict 117 93 8 696 X X X X 56 50 4 6 X X X 23 20 1 996 X X X 98 87 8 1 X X 17 15 1 496 X X 9 86 8 0 X X 511 494 45 8 X 242 234 21 7 X 1155 1079 498 830 128 230 Figure 13 Sample Contrast html Summary 21 The most straightforward use of Contrast is simple differential analysis If a researcher has provided two data sets and one set of criteria as in Figure 9 the resulting Contrast html will have two data set columns One will represent the first sample and the other wil
46. sified proteins These classifications are particularly useful when the option to produce a Bird s Eye view of the proteins is employed see Figure 4 The proteins found in the sample are visualized as small blocks with the colors representing the sequence coverages found for each protein This display can be configured by means of the BirdsEye ini file read from the directory where DTASelect is run see Figure 5 for an example DTASelect can also incorporate data from external programs into its output The aux op tion will cause DTASelect to check the file AuxInfo txt for information pertaining to each locus observed in a sample In this way algorithms for quantifying protein content or providing other data can be integrated with DTASelect A heading for these additional columns will be added to the DTASelect html file and the values entered will appear before the description of each protein For an example of the AuxInfo txt format see Figure 6 Users interested in storing DTASelect s results in a database environment may use its DB option to produce the DB Proteins txt DB Peptides txt and DB Prot2Pep txt files The first two files list the proteins and peptides remaining after DTASelect s filtering with duplicate proteins those that have been grouped to others with identical sequence coverage flagged in the final col umn of the DB Proteins txt file The third shows how which proteins link to which peptides The files are tab deli
47. something like this C IBMJDK java cp C DTASelect DTASelect In this line C IBMJDK java is the filename of IBM s Java Virtual Machine in this case a Windows version is used The cp option ensures that the DTASelect installation directory in this case C DTASelect is on the java classpath DTASelect specifies that this program is run Contrast will specify that Contrast runs instead The ensures that options a user adds to the command line when starting the script are passed to the program for Windows NT and derived operating systems Some virtual machines allow the user to specify the memory available to DTASelect It is helpful to create one script to start DTASelect one for Contrast and a third to start MoreFASTA if used All work from the same classpath but start from different classes within the directory 3 20 Execution Once the program is installed users can start it by running the created script or batch files The program should be run from the directory holding the sequest params file Typically the subdi rectories of spectra and SEQUEST output files are located here f the out files themselves are located in the directory where DTASelect is run rather than in a subdirectory the here or options should be used The options selected by the user should be appended at the end of the line The order of the options does not matter so long as the numerical arguments when required fo
48. tained in the DTASelect results The numbers appearing in this table represent the number of matching peptides for each protein pair minus the number of peptides that don t match If for example protein A has peptides 1 2 and 3 while protein B has peptides 1 2 and 4 there are four peptides matching two copies each of 1 and 2 and two peptides not matching 3 and 4 The similarity score for this pair is 4 2 2 Each protein is listed alongside its similarity to itself which is always equal to two times its number of peptides For users who want to examine how their chromatography is affecting their results DTASe 15 lect provides information about the distribution of peptides among the different chromatographic steps that have been used on each sample and also profiles each step to show when the peptides passing the filters eluted This information found in the DTASelect chroma txt file is presented in tab delimited text when the chroma option is used The first column lists the DAT files These files are the ones from which the dta files spectra are extracted and usually represent a single dimension of chromatography The names are inferred from the common prefixes of the SEQUEST result files The second column shows the total number of peptides after filtering from each DAT file The third fourth and fifth columns show the number of 1 2 and 3 peptides after filtering respectively The remaining columns in each line g
Download Pdf Manuals
Related Search