Getting Started with MSstatsPTM

[Thu Nguyen]

Hello,

I'm trying to give MSstatsPTM a try with my data (which was also TMT-tagged), but I'm having some issues inputting the data. I have attempted to try MSstatsConvert, going from PD to MSstatsTMT (then not sure how to convert to PTM package), and re-analyze the data in MaxQuant so it can go directly to MSstatsPTM but haven't had much luck. I must be missing some important info. If anyone could help me get on the right track, it'll be much appreciated.

1) When doing database searching, was I supposed to search the phospho-enriched fractions separately from the rest of the peptides? I usually search them together for higher confidence in protein ID - but the PTM package mentioned a "PTM" and a "Protein" dataset... I'm a bit unsure of what exactly these are.

2) I found 2 links MSstatsPTM (which has a TMT workflow) and MSstatsTMTPTM... I think I should go with PTM since it has a later release date? just wanted to double-check.

3) Finally (for now), I find that some of these packages work better with R v 4.0 and some require v.4.1. Somehow MSstatsTMT always give me an error saying my annotation and raw input files don't match if I use R ver 4.1 - but if I don't change anything and use R v4.0 instead, it works just fine. Have you noticed something similar? which version of R should I use for PTM?

Thank you so much for your help,

Mi

[Devon Kohler]

Hi Mi,

Hopefully I can clear up a couple of your questions! In terms of going from PD to MSstatsPTM, we do not have a direct converter right now. I have included a brief piece of code that uses the MSstatsTMT converter to put the raw data into MSstatsPTM format. If you have an example phospho PD dataset you can share, I would be happy to work on building a converter directly into MSstatsPTM.

1. The "PTM" and "Protein" datasets should really be two separate MS runs, one phosphorylated and the other not. The non-phosphorylated run is used to adjusted the modified peptides for changes in protein abundance. If you do not have a unmodified MS run, the package can still be run with the "PTM" data only, you just will not be able to do a protein level adjustment..

2. MSstatsPTM is the correct package and can handle TMT experiments now.

3. The versioning is a little different due to Bioconductor's latest release. All packages in Bioconductor v3.12 (the newest version) require R v4.1. I would recommend upgrading your R to version v4.1 and then upgrading the MSstats packages. You should just be able to reinstall MSstatsPTM as it's dependencies include the other packages.

Let me know if you have any more questions!

Best,

Devon

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

# Add site into ProteinName column

raw_ptm_df$ProteinName <- paste(raw_ptm_df$ProteinName, raw_ptm_df$Site, sep = "_")

# Run MSstatsTMT Converters

PTM.data <- MSstatsTMT::PDtoMSstatsTMTFormat(raw_ptm_df, annotation)

PROTEIN.data <- MSstatsTMT::PDtoMSstatsTMTFormat(raw_protein_df,annotation) # If Protein data not available just skip this part

# Combine into one list

raw.input <- list(PTM = PTM.data,

                  PROTEIN = PROTEIN.data) # Use PROTEIN = NULL if protein data not available

[Thu Nguyen]

Hi Devon,

Thank you for your response. It was helpful. I'm trying to work with the codes you gave me and still have 1 (hopefully) final question. What file is raw_ptm_df? --- I tried inputting in both the output from PD (I filtered to only take phosphorylated peptides) and the "Phospho (STY)Sites.txt" from MaxQuant but neither of those files have a "Site" column for me to combine with the ProteinName. 

Thank you again!

Mi