I'm encountering an issue in the MLE process that ...
# crisprseq
i
I'm encountering an issue in the MLE process that seems related to multiprocessing and EOF issues? I've attached the log file as well as my design matrix, and should also note I am starting from a count table. Would appreciate any ideas on how to troubleshoot--thanks!
l
Hi Ittai, could you give me a small sample of your count table too? I’d try to reproduce
And the command line you’re running?
i
Hi Laurence, please find a sample of the count table attached. With respect to the command run, I am using nextflow via an Amazon platform, so I do not run the "nextflow run" command directly, can only see the command run for this task ("mageck mle -k rawcounts.txt -d design_mat.txt -n design_mat"). I am still also inputting the samplesheet CSV (where the sample names are a bit different), but expect that this is basically not read or used at all when I input count data. Let me know if I can provide any other info, thanks for your help!
l
So i managed to run everything with what you provided me
I had this error once with mageck and i recall it was an extra line somewhere or a mal formulated tab
Can you try running it with the snippet that you’ve just sent me and not your whole dataset? If that works it might be stemming from your input count_table
i
Hi Laurence, thanks for looking into it and your prompt reply! I am running on the subset count table now and will keep you apprised. I also have attached the full count table here in case that can help pinpoint the issue. Thanks!
l
ok thanks 🙂 i’m running mageck mle on its own to see if the error pops outside of the pipeline
i
The subset version of the data did successfully finish for me; I took another look at the count table and couldn't discern any malformed delimiters or anything like that...I wonder if it may be due to just the larger scale of the data and increased multiprocessing that must occur as a result?
l
it ran if i only ran mageck mle though, trying with the pipeline now
i
I found the issue was resolved if I changed the MLE process from process_medium to process_high tag, so I think it's safe to say it is a memory issue. Thanks!
l
Hi Ittai, thanks a lot for debugging so much, I’ll add it!! Yay!! 🙂
✅ 1
i
Appreciate your help and prompt responses! 🙂