Hi. 
I'm running into an issue with the TPP/Kojak pipeline. I am running a 
crosslinking analysis using Kojak. As you know, Kojak also reports unlinked 
PSMs. I'm finding that PeptideProphet is filtering out ALL the unlinked 
PSMs in my analysis.
If I process the dataset with Kojak/Percolator I get *8509 unlinked PSMs* 
and 585 crosslinked PSMs with a q value <= 0.05 
Using Kojak/TPP v5.2.1-dev Flammagenitus, Build 202001280920-8004 I get *0 
unlinked PSMs* and 564 crosslinked PSMs with a PeptideProphet error rate of 
>= 0.95
Using Kojak/TPP v5.2.0 Flammagenitus, Build 202001311412-exported I also 
get *0 unlinked PSMs* and 564 crosslinked PSMs with a PeptideProphet error 
rate of >= 0.95 

If I look at the raw data for a specific unlinked PSM, I find things like 
this in the Kojak pep.xml file

spectrum_query 
spectrum="UWPRLumos_2020_0124_AZ_012_AZ866_xlink05.24437.24437.3" 
start_scan="24437" end_scan="24437" precursor_neutral_mass="1782.949616" 
assumed_charge="3" index="25666" retention_time_sec="91.8">
<search_result>
<search_hit hit_rank="1" peptide="KPIDYTILDDIGHGVK" peptide_prev_aa="R" 
peptide_next_aa="V" protein="hAbi2_1-158" protein_link_pos_a="139" 
num_tot_proteins="1" calc_neutral_pep_mass="1782.951515" 
massdiff="0.001899" xlink_type="na" num_tol_term="2" 
num_missed_cleavages="0">
<search_score name="kojak_score" value="5.9750"/>
<search_score name="delta_score" value="5.0900"/>
<search_score name="ppm_error" value="1.0650"/>
<search_score name="e-value" value="3.950e-13"/>
<search_score name="ion_match" value="42"/>
<search_score name="consecutive_ion_match" value="13"/>
<analysis_result analysis="peptideprophet">
<peptideprophet_result probability="0" all_ntt_prob="(0,0,0)" 
analysis="incomplete">
<search_score_summary>
<parameter name="fval" value="7.7950"/>
<parameter name="ntt" value="2"/>
<parameter name="nmc" value="0"/>
<parameter name="massd" value="1.065"/>
<parameter name="isomassd" value="0"/>
</search_score_summary>
</peptideprophet_result>

so scan 24437 identifying KPIDYTILDDIGHGVK gets a e value of 3.950e-13 from 
kojak, which is good, but gets a probability of 0 from peptideprophet (see 
below).

A comet/percolator analysis shows a comet e value of 1.67E-14 and a 
percolator q-value of 0.00009243 for the same PSM (both very good)

A kojak/percolator analysis shows a Kojak e value of  3.950e-13 and a 
percolator q value of  0.000001555 for the same PSM. Again, both very good.

Looking at interact.pep.xml I see:

<spectrum_query 
spectrum="UWPRLumos_2020_0124_AZ_012_AZ866_xlink05.24437.24437.3" 
start_scan="24437" end_scan="24437" precursor_neutral_mass="1782.949616" 
assumed_charge="3" index="25666" retention_time_sec="91.8">
<search_result>
<search_hit hit_rank="1" peptide="KPIDYTILDDIGHGVK" peptide_prev_aa="R" 
peptide_next_aa="V" protein="hAbi2_1-158" protein_link_pos_a="139" 
num_tot_proteins="1" calc_neutral_pep_mass="1782.951515" 
massdiff="0.001899" xlink_type="na" num_tol_term="2" 
num_missed_cleavages="0">
<search_score name="kojak_score" value="5.9750"/>
<search_score name="delta_score" value="5.0900"/>
<search_score name="ppm_error" value="1.0650"/>
<search_score name="e-value" value="3.950e-13"/>
<search_score name="ion_match" value="42"/>
<search_score name="consecutive_ion_match" value="13"/>
<analysis_result analysis="peptideprophet">
<peptideprophet_result probability="0" all_ntt_prob="(0,0,0)" 
analysis="incomplete">
<search_score_summary>
<parameter name="fval" value="7.7950"/>
<parameter name="ntt" value="2"/>
<parameter name="nmc" value="0"/>
<parameter name="massd" value="1.065"/>
<parameter name="isomassd" value="0"/>
</search_score_summary>
</peptideprophet_result>
</analysis_result>
</search_hit>
</search_result>
</spectrum_query>

So I am confused why PeptideProphet is assigning all unlinked peptides in 
these data with a probability of 0. I tried the latest stable version of 
TPP as well as the latest daily build. I see the same results in each. Any 
ideas about this would be very much appreciated. Thanks very much in 
advance!
Alex

-- 
You received this message because you are subscribed to the Google Groups 
"spctools-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/spctools-discuss/52dce033-f9f8-4abf-aeeb-9f30ab5cdaf3%40googlegroups.com.

Reply via email to