[Psidev-pi-dev] SpectraST

[Jennifer S. <jen...@ma...>]

Hi,

Julian and I have been having a look at spectraST and trying to generate 
an example analysisXML document.  Although from what we have done so far 
it seems it is possible to get the basic spectraST details into the file 
with the current schema there are some issues which I will discuss below.

Basically the software takes as input a 'database' of spectra to which 
it matches the query spectrum according to a set of user defined 
parameters and assigns some score - this basic information of course can 
be captured by analysisXML.  Each entry in the search database is 
therefore a peptide and what is actually searched is the spectrum - a 
set of peaks and ion assignments.  Each entry in the database - so each 
peptide/spectrum - has a set of parameters associated with it.  For 
example modifications on the peptide, mass differences between this 
database spectrum and the theoretical spectrum for the peptide and also 
the possibilty of various scores depending on how the spectrum was 
generated.  Some of these details are easy to deal with, for example 
modifications. 

As I understand it modifications maybe set in the user parameters file 
for spectraST which can limit the search - additionally however peptides 
may be identified by spectraST with modifications that were not defined 
in the user parameter file but were present in the search database.  In 
analysisXML this means that the peptide in the SequenceCollection may 
have a modification assigned that is not present in the search 
parameters described in the AnalysisCollection, which I don't think will 
cause any problems.

The interesting part is how the 'database' is generated and we need to 
decide if we think this information is useful.  I downloaded an example 
from NIST and have extracted a very small part of that and pasted it 
below - the start of one database entry (I have not included the entire 
spectra).  You can see from the example that the database contains the 
peptide sequence, various parameters associated with the assignment of 
the peptide to the spectrum and the spectrum itself with the ion 
assignments.  If we were to do a sequence database search and assign 
peptides then in the analysisXML file we would capture the protein 
sequence from the database and the peptide sequence, in the SpectraST 
example the equivalent to the sequence searched is the spectrum.

Before we spend any time producing the example document Julian and I 
thought it would be useful to highlight the issues and discuss in a 
conference call (perhaps tomorrow).

Thanks,

Jenny



-------------------------- example from 
NIST_sigmaups1_v2.0_2008-07-09.msp 
----------------------------------------------------------------

Name: AAAQPHTEIPEGTTAEEAGIGDTPSLEDEAAGHVTQAR/4
MW: 3830.819
Comment: Spec=Consensus Pep=N-Semitryp_insource_confirmed 
Fullname=Q.AAAQPHTEIPEGTTAEEAGIGDTPSLEDEAAGHVTQAR.M/4 Mods=0 
Parent=957.705 Inst=it Mz_diff=0.835 Mz_exact=957.7048 Mz_av=958.263 
Protein="IPI00025499.1|SWISS-PROT:P10636-8|REFSEQ_NP:NP_005901|ENSEMBL:ENSP00000303214 
Tax_Id=9606 Microtubule-associated protein tau isoforM 2" Pseq=92 
Organism="human" 
Se=1^O2:ex=2.005e-011/2.005e-011,td=372925/3.721e+005,pr=1.1e-015/1.1e-015,bs=4.11e-299,b2=4.01e-011,bd=850 
Sample=1/sigma_proof_of_concept_cam,2,2 Nreps=2/2 Missing=0.0901/0.0119 
Parent_med=958.54/0.03 Max2med_orig=136.6/20.3 Dotfull=0.847/0.000 
Dot_cons=0.933/0.019 Unassign_all=0.053 Unassigned=0.047 Dotbest=0.95 
Flags=2,1,0 Naa=38 DUScorr=1.9/1.9/0.38 Dottheory=0.78 Pfin=2.7e+009 
Probcorr=0.82 Tfratio=9.7e+008 Pfract=0
Num peaks: 120
476.3   117     "? 2/2 1.4"
578.5   154     "b17^3/-0.10,b12-46^2/-0.29 2/2 0.4"
579.6   62      "b12-44^2/-0.19,b12-45^2/0.31 2/2 0.0"
643.4   78      "b13-18^2/0.09,b20-46^3/-0.24 2/2 0.1"
649.6   123     "a7/0.27 2/2 1.4"
673.1   95      "? 2/2 1.0"
686.6   73      "b28-46^4/0.03,b28-45^4/-0.22 2/2 0.1"
694.3   181     "b14-18^2/0.47,b14-17^2/-0.03 2/2 1.1"
703.1   106     "b14^2/0.27 2/2 0.1"
706.3   66      "? 2/2 0.3"
711.8   101     "y6/0.41 2/2 0.2"
729.5   291     "b15-18^2/0.15,b15-17^2/-0.35 2/2 0.3"
738.5   255     "b15^2/0.15 2/2 1.1"
744.1   93      "b23-18^3/0.09,b23-17^3/-0.24 2/2 0.4"
746.8   76      "? 2/2 0.4"
761.7   58      "b31-17^4/0.10,b31-18^4/0.35 2/1 0.2"

-----------------------------------------------------------------------------------------------------------