SpectronauttoMSstatsPTMFormat problems PeptideSequence

[Ryan Trouvé]

Hello

I am trying to anlayse my phospho-data, but I already have issues in the importation from Spectronaut (v21.0) with SpectronauttoMSstatsPTMFormat. I had no issues with the normal proteomic analysis and its uploading with SpectronauttoMSstatsFormat.

With running following code: 

SpectronauttoMSstatsPTMFormat(
  "20261002_PCF000732_phospho_RT_Report.tsv",
  annotation = "Phosphoproteomics metadata.csv",
  fasta_path = "UP000000589_10090.fasta",
  protein_input = "20261002_PCF000732_RT_Report.tsv",
  annotation_protein = "Proteomics metadata.csv",
  use_unmod_peptides = FALSE, 
  intensity = "PeakArea",
  mod_id = "\\[Phospho \\(STY\\)\\]",
  fasta_protein_name = "uniprot_iso",
  remove_other_mods = TRUE,
  filter_with_Qvalue = TRUE,
  qvalue_cutoff = 0.01,
  useUniquePeptide = TRUE,
  removeFewMeasurements = TRUE,
  removeProtein_with1Feature = FALSE,
  summaryforMultipleRows = max,
  use_log_file = TRUE,
  append = FALSE,
  verbose = TRUE,
  log_file_path = NULL
)

I always get the following error:

Error in `[.data.table`(data, , c(protein_name_col, unmod_pep_col, mod_pep_col), : columns not found: [PG.ProteinGroups, PeptideSequence, EG.PrecursorId]

However, Spectronaut does not have any columns for 'PeptideSequence', and I have loaded the PEP.StrippedSequence and EG.ModifiedSequence columns:

> colnames(phospho) [1] "R.Condition" "R.FileName" "R.Replicate" [4] "PG.Genes" "PG.ProteinAccessions" "PG.ProteinGroups" [7] "PG.IsSingleHit" "PG.Qvalue" "PG.Quantity" [10] "PEP.GroupingKey" "PEP.IsProteotypic" "PEP.PeptidePosition" [13] "PEP.StrippedSequence" "PEP.Quantity" "EG.iRTPredicted" [16] "EG.IsDecoy" "EG.Library" "EG.PrecursorId" [19] "EG.ModifiedSequence" "EG.Qvalue" "EG.ProteinPTMLocations" [22] "EG.PTMPositions [Phospho (STY)]" "EG.PTMSites [Phospho (STY)]" "EG.PTMAssayProbability" [25] "EG.PTMLocalizationProbabilities" "FG.Charge" "FG.Id" [28] "FG.PrecMz" "FG.Quantity" "F.Charge" [31] "F.FrgIon" "F.FrgLossType" "F.FrgMz" [34] "F.FrgNum" "F.FrgType" "F.PredictedRelativeIntensity" [37] "F.ExcludedFromQuantification" "F.NormalizedPeakArea" "F.NormalizedPeakHeight" [40] "F.PeakArea" "F.PeakHeight"

I now just read the raw phospho tsv with data.table::fread, and have added a column 'PeptideSequence' based on the excisting column 'PEP.StrippedSequence'. This seems to work for the uploading, but I wanted to make sure that this would 1) not effect downstream analysis and 2) that you were aware this occurs.

Thank you!

Best

Ryan

[Anthony Wu]

Hi Ryan,

Thanks for pointing this out.  I did an investigation and this looks like it's a misleading error message.  SpectronauttoMSstatsPTMFormat should work as long as you have the column EG.PrecursorId or EG.ModifiedSequence (it looks like both of those columns exist for you).

I actually think the main issue is that the input for SpectronauttoMSstatsPTMFormat needs to be a data.frame rather than a filename (so doing SpectronauttoMSstatsPTMFormat(

    "20261002_PCF000732_phospho_RT_Report.tsv", .... ) doesn't work)

We'll first look to fix the misleading error message, then update the docs so that it's clear the input to the function is a dataframe.  

Tony