Open Chuang1118 opened 2 years ago
I think the read in and count kmer is not the sophisticated part, and should not crash normally. I feel this is more like a system error. Did you run TRUST4 on a cluster and when rerun the cmd is it on the same node?
I use TRUST4 by building singularity container and launch job by Snakemake, because I've also gene expression part. Now, I am in stage define parameters hardware in multiple cores server. I have tested 2 servers, I counter this problem 1 time for human, 1 time for mouse. This error is not reproducible. Run and reRun are in same server. I plan to run TRUST4 in HPC Slurm, since my lab production many BCR data.
I'm not familiar with the containers, and there could be some issues with virtualization. Please let me know if it still crashes running on slurm without singularity.
Hello trust4,
Version TRUST4 use Output all the required fields for AIRR format https://github.com/liulab-dfci/TRUST4/pull/138
CMD
BUG: In time point [Wed Sep 28 22:22:50 2022] is blank.
[Wed Sep 28 22:16:19 2022] Read in and count kmers for 11900000 reads. [Wed Sep 28 22:16:23 2022] Read in and count kmers for 12000000 reads. [Wed Sep 28 22:16:26 2022] Read in and count kmers for 12100000 reads. [Wed Sep 28 22:22:50 2022]
cmd observed in server & I have enough RAM & Mem
reRUN same CMD, this time is work fine. [Thu Sep 29 12:07:01 2022] Read in and count kmers for 11900000 reads. [Thu Sep 29 12:07:04 2022] Read in and count kmers for 12000000 reads. [Thu Sep 29 12:07:07 2022] Read in and count kmers for 12100000 reads. [Thu Sep 29 12:13:47 2022] Found 12156257 reads. [Thu Sep 29 12:14:12 2022] Finish sorting the reads.
I counter this issue both in ref mouse & human.
Please advise me,
Chuang