Error filtering iProphet output from multiple search engines

[ohickl]

Hi,

I wanted to combine the outputs of multiple search engines from the same input, but I am getting:

> ${phil_bin} filter --sequential \
>                    --razor \
>                    --picked \
>                    --pepProb ${pep_thresh} \
>                    --protProb ${prot_thresh} \
>                    --tag ${dec_prefix} \
>                    ${filter_params} \
>                    --pepxml ${pepxml} \
>                    --protxml ${protxml}
INFO[13:07:46] Executing Filter  v4.8.1
INFO[13:07:46] Processing peptide identification files
INFO[13:07:46] Parsing se_mvh_wdp_xcorr_msf_iprophet.pep.xml
panic: nil

goroutine 12 [running]:
philosopher/lib/id.(*PepXML).Read(0xc000252420, {0x7ffc0aca38e4, 0x25})
        /workspace/philosopher/lib/id/pep.go:210 +0xf4e
philosopher/lib/id.ReadPepXMLInput.func1(0x0, {0x7ffc0aca38e4, 0x25})
        /workspace/philosopher/lib/id/pep.go:330 +0x105
philosopher/lib/id.ReadPepXMLInput.func2(0x0?, {0x7ffc0aca38e4, 0x25})
        /workspace/philosopher/lib/id/pep.go:366 +0xa2
created by philosopher/lib/id.ReadPepXMLInput
        /workspace/philosopher/lib/id/pep.go:363 +0x998

Ther general workflow:

# Run PeptideProphet
for pepxml in ${pep_xmls[@]}; do
  type=${pepxml##*${sample}/}
  type=${type%%/*}

  type_alt=${pepxml%%/${sample}*}
  type_alt=${type_alt##*/}
  
  if [[ "${type_alt}" == "msfragger" ]]; then
    type="msf"
  elif [[ "${type_alt}" == "comet" ]]; then
    type="com"
  fi

  ln ${pepxml} ${work_dir}/${type}.pep.xml

  ${phil_bin} peptideprophet --database ${db_raw} \
                             --nonparam \
                             --decoyprobs \
                             --decoy ${dec_prefix} \
                             --minprob ${pep_thresh} \
                             --output interact_pep_thresh_${pep_thresh} \
                             ${pep_proph_params} \
                             ${work_dir}/${type}.pep.xml > ${pep_p_log} 2>&1
done

# Run iProphet
${phil_bin} iprophet --threads ${threads} \
                     --decoy ${dec_prefix} \
                     --output se_mvh_wdp_xcorr_msf_iprophet \
                      interact_pep_thresh_${pep_thresh}*pep.xml > ${i_p_log} 2>&1

# Run ProteinProphet
if [[ ${mode} == "closed" ]]; then
  prot_proph_params=""
elif [[ ${mode} == "open" ]]; then
  prot_proph_params="--maxppmdiff 2000000"
fi

echo "Running ProteinProphet."
pepxml="se_mvh_wdp_xcorr_msf_iprophet.pep.xml"

${phil_bin} proteinprophet ${prot_proph_params} \
                           --minprob ${pep_thresh} \
                           --iprophet \
                           --output "interact_pep_thresh_${pep_thresh}" \
                           ${pepxml} > ${prot_p_log} 2>&1

protxml=$(ls interact_pep_thresh_${pep_thresh}*.prot.xml)

I assume, this is because, in addition to msfragger and comet, I am using input from another tool, which seems to output a variant pep.xml format. Is it possible to locate and fix the incompatibility? Only combining msfragger and comet runs fine.

I have attached a truncated example pep.xml

Best

Oskar

example-mvh.pep.zip

[anesvi]

Philosopher only supports MSFragger and Comet. We do not have plans to support other search engines at this time. Alexey

[ohickl]

Thanks for the swift reply!
Ok, I didnt mean it as a feature request. Im fine with trying to make it work myself, if possible. But since i am not too familiar with pep.xml, I am seeking some guidance on what to look for, since I assumed it would be easy to spot for someone with more experience.

[anesvi]

I am afraid it would be a lot of work involved. Philosopher needs to read specific search engine scores and print them in the summary files. Perhaps Felipe Leprevost will be able to respond to you and provide some guidance.