can't finish the process

[RiMaxim]

Hi,

I have some problems with MANTIS. I can't finish the process. I used WGS data (70 Gb - normal and 150 Gb - tumor) and 3 mln rows of loci bed basing on hg38 genome.

Also I used only chr16 from bam files (~2-3 Gb) and loci.bed for chr16 (100K) lines, but my process can't finish.

Though?

Microsatellite Analysis for Normal-Tumor InStability (v1.0.4)
/usr/bin/python3 /home/ubuntu/1/MANTIS/kmer_repeat_counter.py
-b /home/ubuntu/1/MANTIS/loci.bed2
-n /home/ubuntu/1/MANTIS/DNASeq_blood.sorted.chr16.bam
-t /home/ubuntu/1/MANTIS/DNASeq_cancer.sorted.chr16.bam
-o /home/ubuntu/1/MANTIS/1.kmer_counts.txt
--min-read-quality 25.0
--min-locus-quality 30.0
--min-read-length 35
--genome /home/ubuntu/1/MANTIS/hg38.fa
--threads 64
Getting repeat counts for repeat units (k-mers) ...

Regards, Maxim

[rbonneville]

Hello Maxim,

I would recommend using fewer threads (try 3-5 at first), along with fewer loci. In our landscape analysis of TCGA data we used 2,530 loci (https://ascopubs.org/doi/suppl/10.1200/PO.17.00073), so 100,000 loci is overkill for MSI calling.

[RiMaxim]

Hi I used 65 threads for 10 loci. And i have problem.

[rbonneville]

Try substantially reducing your thread count; even if you have sufficient CPU cores this many threads will probably be I/O limited.

[ghosholivia]

Hi @rbonneville ,
I'm having the same issue. I had tried with thread 5 but it's been 4 hrs and it is still running.
The loci.bed I'm using is of 20lakh lines. Is that a problem?
Can you guide me please?

[RiMaxim]

Hi, I used 1 core and I solved my problem. It seems to be very strange. Regards, Maxim ср, 4 нояб. 2020 г. в 19:20, ghosholivia <[email protected]>:

…

Hi @rbonneville <https://github.com/rbonneville> , I'm having the same issue. I had tried with thread 5 but it's been 4 hrs and it is still running. Can you guide me please? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#57 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADZF6Y5N6E6ADP3MOXTC5TDSOFBKDANCNFSM4SFQAXBQ> .

[ghosholivia]

Hi, I used 1 core and I solved my problem. It seems to be very strange. Regards, Maxim ср, 4 нояб. 2020 г. в 19:20, ghosholivia [email protected]:
…
Hi @rbonneville https://github.com/rbonneville , I'm having the same issue. I had tried with thread 5 but it's been 4 hrs and it is still running. Can you guide me please? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#57 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADZF6Y5N6E6ADP3MOXTC5TDSOFBKDANCNFSM4SFQAXBQ .

Yeah. It ran successfully but did it take 1 day for WGS bam

[rbonneville]

This seems like a potential threading issue. If you notice stalling with many threads, could you please terminate the process and post a traceback?

[ghosholivia]

Yes, it ran successfully with low threads.

Thanks.