SIFpackets User Guide
Contents
1. 6 peaks per amino acids A lower value will reduce the number of peaks other than b and y inside the spectrum reducing the effectiveness of PSA increasing it will lead to keep lot of noise and thus to a much slower method e The threshold PP parameter is the minimal score a packet must obtain to consider an alignment position The default value 8 00 means that at least a b ion or a y ion must be used in the alignment to consider it A high value will lead to miss lots of positions and thus to a bad scoring alignment and a low value will only keep lot of wrong PP e The window width to filter PP parameter is the width of the window used to filter the PP The default value 110 matches the average size of an amino acid e The window quantity to filter PP parameter is the number of PP to keeps for each position of the sliding window By default 6 PP are kept for each position To many positions will strongly slower the method and to few will speed it up but will also decrease quality This parameter strongly relies on the quality of the experimental spectra and on having a model of packet perfectly adapted to the data The window has a batch list at the bottom left side When a task is correctly parameter it must be added to the Configuration to run list with the add button Several task may be added The remove button will delete the selected task from this list The launch comparison button will starting processing the tasks from thi2. induce a fixed amount of modification per peptide using an equal probability to each amino acids to mutate into another ran domly chosen amino acid either induce mutation using PAM matrices as defined by Dayhoff The value of this fields can be either an integer being the exact number of modification added in each peptide or a string representing the PAM matrix to use e g PAM10 is a correct value The default value 0 means that no modification are applied to the peptides The field Filter can be used to filter proteins using their names Only those with the given string will be kept It could be useful in some cases in order to keep only one organism or even one chromosome from a databank An empty field means that no filter is applied The field Output File Name allow to enter the name used to store the prepared databank A name is suggested by default it consists of the FASTA file name concatenated with the modifications applied The Digest button start the preparation process with the chosen pa rameters A progression bar allows to follow the process The lower part of the window allow to look at the databank that were previously prepared Selecting one in the Existing Databank list allow to look at the parameters used to create it and at the number of peptides composing it 3 2 Comparison of Spectra against a Databank The comparison method PacketSpectralAlignment is available through the menubar Tools Se
3. SIFpackets User Guide Freddy CLIQUET LINA UMR CNRS 6241 Universit de Nantes 2 rue de la Houssinire 44322 Nantes Cedex 03 France freddy cliquet univ nantes fr Guillaume FERTIN LINA UMR CNRS 6241 Universit de Nantes 2 rue de la Houssinire 44322 Nantes Cedex 03 France guillaume fertin univ nantes fr Irena RUSU LINA UMR CNRS 6241 Universit de Nantes 2 rue de la Houssinire 44322 Nantes Cedex 03 France irena rusu univ nantes fr Dominique TESSIER UR1268 BIA INRA Rue de la Graudire BP 71627 44316 Nantes France dominique tessier nantes inra fr April 28 2010 Contents 1 Introduction 2 Installation 3 Functionality 3 1 Databank preparation 3 2 Comparison of Spectra against a Databank 3 3 Analyzing the results 3 3 1 From peptides to Proteins 2 05 33 amp Heb Sa EE aS 3 3 2 Area Under the ROC Curve 3 3 3 Visualizing spectra 4 Parameters References 11 11 1 Introduction The SIFpackets application can perform comparisons between experimental tandem mass spectra MS MS spectra and proteins databank This compar ison is done using the algorithm PacketSpectralAlignment PSA developed by Cliquet et al in 1 and the framework detailed in 2 PSA allows to identify in spite of modifications insertion suppression or substitution of one or several amino acids This is especially useful to search homologous proteins in related organisms 2 Installat
4. arch Spectra This is the tool that will compare all the spectra from an experimental dataset with all the candidates peptides from the chosen databank Several options are available here e The Spectra File can be chosen in a list All recognized spectra files must be stored in the following directory lt Application Location gt data databank fasta The application supports now two different input format the first one is the Mascot Generic Format MGF and the second one is the XML format from ProteinLynz e The Databank File list allows to select the databank to search on These databanks have been created using the Databank preparation tool subsection 3 1 e The field Output file name contains the filename used to store the results files A default name composed of a concatenation of the name with the databank file name is proposed e The field Number of Core allows to set the number of core to use on the computer thus allowing the creation of the same number of thread The default value correspond to the total number of threads of the current computer This value must be lower if it is needed to run other application while searching spectra or if the computer has memory problems e The K parameter corresponds to the number of modifications tolerated per 100 Daltons The default value 0 15 corresponds to an average of one modification per 6 amino acids e The Delta parameter represents the maximal percentage of tolerated mass d
5. articular cases the visualization tool can compute the Area Under the ROC Curve AUC a value that attest the quality of the results This is only possible if the good association peptide spectrum is known for for each spectrum of the dataset It is possible to do so if a peptide prophet result file exists and is place with the spectra file Right now it only works for spectra in the MGF file format 3 3 3 Visualizing spectra When an element from the peptide list is selected it is possible to visualize the experimental spectrum and the theoretical one together to see how they match This functionality is particularly interesting in presence of modifica tions especially as it has been designed to allow the live modification of the peptide to adjust it on the experimental spectrum The visualization window contains sevral elements of interest e The first one is the list of modifications that allow the better match according to PSA Each element of this list correspond to one mod ification and contains at first the localization of the modification in Daltons not the number of the amino acid then the new mass of the element found in this place and then its previous mass Using those two masses it is possible to know at least to have an idea of what was the element and more importantly of what it could be after the modification e Then a field with the matching unmodified theoretical peptide The peptide from this field can be modi
6. e positions used by PSA The blue ones represents positions where packets are effectively aligned The higher a possible position is in this graph the better the score of the alignment will be e Finally a reminder of the alignment score is shown at the bottom of the screen this score is updated with the change of the peptide in the field 4 Parameters Default parameters are frequently suggested they have been fixed using sev eral experimentation on real sets of data coming from a Quadrupole Time of flight Q TOF mass spectrometer They must give good results but may eventually depend on the mass spectrometer you use In addition right now the only model of packet used to align spectra relies on Q TOF data References 1 F Cliquet G Fertin I Rusu and D Tessier Comparison of spectra in unsequenced species In Proc 4th Brazilian Symposium on Bioinformatics BSB 2009 volume 5676 pages 24 35 2009 2 F Cliquet G Fertin I Rusu and D Tessier Proper alignment of ms ms spectra from unsequenced species In BIOCOMP 2010 11
7. fied and the spectrum updated using the Refresh Spectrum button This field supports a particular format All peptides coded using the one letter code of each amino acids is supported An unknown element of known mass can be created using the notation _x_ where x is its mass in Daltons This represents an element so a fragmentation is expected immediately before and after it The mass of an amino acid can be changed by adding 2 or x just after it to add or subtract x Daltons For example in AM 16GV the 16 will increase the mass of the methionine M of 16 Daltons corresponding to an oxidation e Several spectra representations are then proposed The blue peaks represent N terminal ions and the red ones represent C terminal ions The first spectrum represents the theoretical spectrum from the peptide entered in the field just above The second one represents the experimental spectrum after ap plication of symmetry thus there is a duplication of the peaks 10 The third one represents the experimental spectrum with in green the peaks that are aligned with the theoretical spectra according to PSA Be careful symmetry has been applied on this spectrum thus all blue peaks have a red duplicate one in most of the cases only one of them will be aligned Considering this not having all the high intensity peaks aligned could be normal The last representation is a plot of the possibl
8. ifference between the modified and unmodified peptide The default value 20 means that if there is more than 20 mass difference between an experimental peptide given by the parent mass of the experimental spectrum and a databank peptide the comparison is not done and the similarity score is assumed to be 0 The higher the Delta value is the slower the method will be This value is a compromise between speed and quality and highly depend on the average number of modifications expected Our method apply several preprocessing method on the experimental spectra At first a filtering on the peaks to eliminate noise by sliding a window on the spectra and keeping only the most intense peaks for each positions of this window Then the method compute what we have called possible positions PP These are position inside a spectrum that will prob ably lead to a high scoring alignment with a packet These position are filter at first with a threshold the minimal score given by a position and then by a sliding window to keep only the most promising in a certain area of the spectra It is possible to change all of these parameters e The filter window width corresponds to the width of the window used to filter the peaks The default value 110 matches the average size of an amino acid e The filter window quantity parameter is the number of peaks kept for each position of the window The default value 6 means that we expect to have on average
9. ion The installation folder of PacketSpectralAlignment PSA contain several di rectories to store all the data needed and the output files e data contains all the data used by the application databank contains the databank to which the spectra will be com pared both fasta and prepared for PSA databanks x fasta contains the fasta of the databanks matrices contains the PAM matrix to allow the creation of modifications inside databanks matrices contains precomputed alignment used for the AUC com putation output contains all the files coming from the comparisons of the spectra R contains special files created in order to compute ROC curve using the R language spectrum contains the experimental spectra 3 Functionality 3 1 Databank preparation The Databank preparation functionality is necessary in order to compare ex perimental MS MS Spectra to a protein databank This is here that the data bank will be virtually digested by an enzyme Currently only the trypsin 3 can be used The FASTA file of the databank to use is needed and must be place directly in the appropriate folder lt Application Location gt data databank fasta This tool can be access in the menubar Tools Databank preparation Fasta File 18mix KA reload Min mass in Da 600 Max mass in Da 3000 Modifications 1 Filter Ouput File Name 18mix 1 Waiting for exec
10. nt peptides identified in the protein In additon when a protein is selected the Peptides list shows all the couples peptide spectrum used to identify the protein Each couple when selected shows in the lower part of the window some informations e The Peptide id is the id of the peptide in the databank file e The Spectrum id is the id of the spectrum in the result file e The PSA score is the score given by the comparison method PSA e The peptide sequence is the peptide sequence that matches the best with the spectrum e The Is low complexity boolean field gives a complexity information on the peptide sequence If the sequence contains a string repetition of Glycine it is consider as low complexity and thus will have lower the peptide score e The peptide score is the contribution to the score of the protein for this couple spectrum peptide It is computed using several factor such as the length of the peptide sequence and its complexity e The Mass difference field gives the difference of mass between the ex perimental spectrum and the theoretical one If there are no modifi cation this difference should be close to 0 it is possible to see a very small variation due to the mass spectrometer precision If there are modifications this mass difference can give informations on the modifi cation much more useful information are available see the Visualizing spectra part 3 3 3 3 3 2 Area Under the ROC Curve In some very p
11. s list top to bottom showing the progression with the progression bar On the bottom right side is the Existing results list to look at already existing files 3 3 Analyzing the results The main window of the application allows to analyze results of the search comparison It is possible to load the existing result file by selecting it in the Results to visualize list if the element does not appear in the list it is possible to reload the list and clicking on the Visualize Results button Loading results may take some time since the method will try to identify proteins based on the already identified peptides On the upper part of the screen will be display some data concerning the analysis such as the spectrum file and databank file used or the executions parameters Then this tool comports several possibilities identification of proteins computing Area under the ROC Curve and Visualizing Spectra 3 3 1 From peptides to Proteins All the proteins identified by at least one identified peptide from the search are listed in the Proteins list Each protein has an associated score the score is between 0 and 100 the higher this score is the more chances there is that the protein is correctly identified Selecting a protein gives its full name Protein name field and its score with some details score complete that depends on the number of pf spectra used to identify the protein and score distinct that depends on the number of differe
12. ution Existing Databank 18mix_rice PAM20 an Fasta file 18mix Modification 0 Mass range 600 to 3000 Number of peptides 454 Filter Figure 1 Databank Preparation window In the Databank Preparation Window Figure 1 there are several options available They are listed and explained here e The Fasta File list allows to chose the databank FASTA file files ap pears in the list without the extension fasta It is possible to reload the content of the list in the case that a new file have been copied in the fasta directory by clicking on the reload button e The Min mass et Max mass fields allow to select the range within we will conserve the peptides after the digestion This mass usually depends on the tolerance of the mass spectrometer used The default 4 value min 600 Da and max 3000 Da are common values for Q TOF mass spectrometers wrong estimation of these two parameters can cause some problems if they are too restrictive high Min mass and low Max mass the application may miss expected peptides and on the contrary giving a low Min mass value or a high Max mass will allow too much comparison and no good results are expected from these additional peptides The field Modifications allows to modify the peptides during the diges tion This possibility is for a purpose of tests to verify the ability of the method to identify peptides in spite of modifications The method can either
Download Pdf Manuals
Related Search