Issue running MANTIS

[chloe35430]

Hello,

I have an issue while running MANTIS on exome datasets.
I wrote the following code lines :

python2.7 mantis.py -b ../../Test_MANTIS/genome_scan.bed --genome ../../Test_MSISensor/IndexedGenome/hg19.fa -n ../../Test_MANTIS/constit.cleaned.bam -t ../../Test_MANTIS/somatic.cleaned.bam -o ../../Test_MANTIS/resultats_mantis_hg19.txt -mrq 20.0 -mlq 25.0 -mlc 20 -mrr 1

And I got these printed on my terminal:

Microsatellite Analysis for Normal-Tumor InStability (v1.0.4)
python /Users/chloe/work/MANTIS/kmer_repeat_counter.py
-b /Users/chloe/Test_MANTIS/genome_scan.bed
-n /Users/chloe/Test_MANTIS/albert_constit.cleaned.bam
-t /Users/chloe/Test_MANTIS/albert_somatic.cleaned.bam
-o /Users/chloe/Test_MANTIS/resultats_mantis_albert_hg19.kmer_counts.txt
--min-read-quality 20.0
--min-locus-quality 25.0
--min-read-length 35
--genome /Users/chloe/Test_MSISensor/IndexedGenome/hg19.fa
--threads 1
Getting repeat counts for repeat units (k-mers) ...

Then, after 48h running, I got this error that I don't understand :

Traceback (most recent call last):
File "/Users/chloe/work/MANTIS/kmer_repeat_counter.py", line 866, in
normal = krc.process(config['normal_filepath'], msi_loci, config)
File "/Users/chloe/work/MANTIS/kmer_repeat_counter.py", line 614, in process
self.status_check(queue_out.qsize())
File "/System/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/multiprocessing/queues.py", line 143, in qsize
return self._maxsize - self._sem._semlock._get_value()
NotImplementedError

I don't get where this error comes from. Could you help me please?

Thanks,
Chloé

[rbonneville]

Hello @chloe35430, could you please share your genome BED file?

[chloe35430]

Hello, @rbonneville, thanks for your quick answer !

It is a whole genome bed file, it weighs 1.15 Go and the .bed is not supported.

Here is the header of the file:
chr1 10485 10498 (GCCC)3 0 +
chr1 10629 10635 (GC)3 0 +
chr1 10652 10658 (AG)3 0 +
chr1 10658 10664 (GC)3 0 +
chr1 10681 10687 (AG)3 0 +
chr1 10687 10693 (GC)3 0 +
chr1 10710 10716 (AG)3 0 +
chr1 10716 10722 (GC)3 0 +
chr1 10739 10745 (AG)3 0 +
chr1 10745 10751 (GC)3 0 +

I used Repeat Finder to create it and it seemed to go without any issue...
Let me know if you need anything else.

Thanks for the help !

Chloé

[rbonneville]

@chloe35430, MANTIS is crashing because of this extremely large number of microsatellite loci. RepeatFinder generates a .bed file of all putative microsatellites in the supplied genome. This should be filtered to only use microsatellites within your capture region or other region of interest.

[chloe35430]

@rbonneville , sorry for the late answer.
I filtered the file generated by RepeatFinder to get only the loci sequenced in the experiment. My bedfile weighs now 25 ko, but I still get the same error "NotImplemented Error", just much faster than before.

Do you have any idea on how to solve this ?

Thanks for your help,

[audyavar]

For exomeSeq data, how do you go about filtering the BED File generated by RepeatFinder? I have paired end exomeseq data from primary tumors. Do you need to filter the loci for exomeseq data?

[rbonneville]

@chloe35430 how many lines are in your BED file?

[rbonneville]

@audyavar The BED file from RepeatFinder should be filtered for loci included in your capture region. This can be done easily with bedtools, for example:

bedtools intersect -a microsatellites.bed -b regions.bed -wa > microsatellites_filtered.bed

[audyavar]

Thanks. But isn’t whole exomeseq covering the entire exome? Is filtering needed there? If so then what regions should I filter for whole exomeseq seq data? This is very new to me so any help would be appreciated. Best Akshata

…

________________________________ From: rbonneville <[email protected]> Sent: Thursday, April 18, 2019 7:43 AM To: OSU-SRLab/MANTIS Cc: Akshata Udyavar; Mention Subject: Re: [OSU-SRLab/MANTIS] Issue running MANTIS (#42) @audyavar<https://github.com/audyavar> The BED file from RepeatFinder should be filtered for loci included in your capture region. This can be done easily with bedtools, for example: bedtools intersect -a microsatellites.bed -b regions.bed -wa > microsatellites_filtered.bed — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub<#42 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AASBJUOT4CNKSCNHYC4YQ6TPRCCHLANCNFSM4HB7SMCQ>.

[chloe35430]

@rbonneville My bedfile counts 828 lines. Is it still too many lines ?

[rbonneville]

@chloe35430 That is definitely not too many lines. I believe you are encountering a Mac-specific issue with Python multithreading, in particular because MANTIS uses semaphores which are not properly implemented in OS X.

In kmer_repeat_counter.py, try replacing BoundedSemaphore with Semaphore. This will break thread safety, but since you're only using one thread anyway it should be less of an issue.

[rbonneville]

@audyavar Your whole exome capture kit should have come with a list of target regions, likely BED formatted. It is necessary to filter for the exome because RepeatFinder outputs putative microsatellites throughout the entire genome.

[chloe35430]

@rbonneville I just tried the modification you suggested, it still doesn't work and produces the same error. Do you have any other idea to solve this ?
I am going to try running it under a linux system, to see if I get the same issue, but I need to get it to work on Mac too...