dataProcess unable to complete

49 views

Skip to first unread message

Scott Lyons

unread,

Oct 11, 2022, 3:49:31 PM10/11/22

to MSstats

I've been using the MSstats.csv output from fragpipe for my unlabeled DDA MSstats analysis for a while now, but for some reason with my most recent data set the dataProcess function just seems to stall no matter how long I leave it. When I say stalls I mean on the 'Start the summarization per subplot...' step it doesn't get past 1%. However, the task manager still shows that my R session is actively using CPU and memory, so I know it's not frozen. If I take a smaller subset of the data (100,000 features) it stalls at the same place but at 6%. I've troubleshooted by restarting R, my PC, updating all packages, checking that I have the most recent MSstats version, and re-running the analysis on a similar data set (which worked as expected). I attached the log file and a sample of the fragpipe output data. I'm wondering If i overlooked something obvious?

I appreciate the help,

Scott

MSstats_dataProcess_log_2022_10_11_15_17_15.log

subset_fragpipe_output.csv

Scott Lyons

unread,

Oct 12, 2022, 10:45:44 AM10/12/22

to MSstats

Okay So the issue it turns out was RAM. After looking around on the forms I found the one "MSstats::dataProcess RAM overflow for SpectroNaut input" very useful. The difficult part was that I didn't get any error telling me what the possible issue could be. After following Christian Schori suggestion of setting the R_MAX_VSIZE to a higher value, with no luck, I followed his later suggestion of setting the "featureSubset" argument in dataProcess to "topN", then setting "n_top_feature" = 1000. The processing step looked like it initially stalled but after a few minutes was able to finish.