Understanding the MSstatsPTM input

[Joaquin Reyna]

Hi MSstatsPTM team,

Thanks for building MSstatsPTM. I was recently asked to use this as part of a collaboration with my proteomics core, so please forgive me if I miss any details.

As I've been going through this package I have found it difficult to understand the underlying input data structure. At a high level, I understand this software/statistical framework is trying to correct for protein-level changes by integrating data from two experiments:

1) Global protein data
2) Phosphopeptide data

As such, I was imagining the input to MSstatsPTM would be either two separate files, one for each experiment, or a single file that concatenates these two datasets together. Is this correct?

I don't see any vignette mention this, which makes it difficult to understand how to get started. I'm wondering if you could provide more documentation on the expected input format, either through additional vignettes or a detailed GitHub README.

Many thanks,
Joaquin

[Anthony Wu]

Hi Joaquin,

Apologies for delays as we recently wrapped up our May Institute short course.

You can found our vignettes on our Github repository here.  

Specifically, I would focus on the converters.  For example, with Fragpipe converter, you can see that the first and second parameters refer to the phosphopeptide experiment's data and experimental design respectively, and the third and fourth parameters refer to the global profiling experiment's data and experimental design respectively.  Other converters will have similar patterns, which can be found in their manuals (e.g. on Rstudio, run `?MaxQtoMSstatsPTMFormat` to see which parameters correspond to the phosphopeptide vs global profiling data for the MaxQuant converter).

Which spectral processing tool are you using?  I can give you more specific pointers on which parameters represent the phosphopeptide vs global profiling experiment based on the spectral processing tool converter.

Thanks,

Tony

[Joaquin Reyna]

Thanks Tony,

My team is working with Spectronaut. We recently discovered that Spectronaut generates MSstats output. This took us some time to discover and it would be helpful for newcomers if this were somewhere in the vignettes you mentioned.

Best, 

Joaquin

[Anthony Wu]

Hey Joaquin,

For the Spectronaut converter, input and annotation parameters refers to the phosphopeptide's experiment's data and experimental design.  Then input_protein and annotation_protein refers to the global profiling experiment's data and experimental design. 

I agree the vignettes need to be updated.  I'm noticing most of the converters in the vignette I linked earlier don't have a sample global profiling proteomic dataset for adjustment (only Fragpipe has it), which makes it difficult to figure out how to incorporate the global profiling adjustment.  We'll update the vignettes to include a sample global profiling dataset for the other converters for the next release.

Tony

[Joaquin Reyna]

Sounds good! I'll look through this and see if I can get it going. Currently I'm squeezed on time but I'll jot down some notes once I get things going and I can forward them to y'all if that helps the vignette.

Best,

Joaquin