10971 Segmentation fault (core dumped)

[tamaraprieto]

Hi,

I ran SCIPhI for 200000 iterations and at the end, I got a segmentation fault. I am still running version 0.1.2. Do you think it could be a bug solved in 0.1.3 and I should upgrade or do you suspect of any other reason?

... iterations: 190000 score: 0 bestScore: 0 num trees: 1 num Samples: 30 total # mut: 0 currently used: 0 normal - freq: 0.294349812831224 tmp: 0.290966440888769 SD: 0.0141906754859326 count: 568 trails: 568 normal - overDis: 549.01817943582 tmp: 535.818107369113 SD: 7.09533774296629 count: 493 trails: 493 normla - alpha: 161.603398357873 beta: 387.414781077947 mutation - overDis: 15.2505254570906 tmp: 15.180311698458 SD: 0.141906754859326 count: 649 trails: 649 mutation - alpha: 4.63260318596367 beta: 12.1142520424177 drop: 0.633748158249681 SD: 0.0141906754859326 count: 556 trails: 556 lambda: 0.533507173081417 SD: 0.0141906754859326 count: 502 trails: 502 iterations: 0 score: 0 bestScore: 0 num trees: 1 /var/log/slurm/spool_slurmd/job2059288/slurm_script: line 58: 10971 Segmentation fault (core dumped) sciphi -i ${RESDIR}/${REF}.fa --in ${WORKDIR}/SciphiSamples.${PATIENT}.txt -o Sciphi -r 2 -l 100000 --lz 1 --ls 100000 --pr 0.0001 --seed ${oddnumber} ${WORKDIR}/Sciphi.${PATIENT}.mpileup

Thanks in advance,

Tamara

[singerj]

Dear Tamara,

The problem is that during candidate loci identification not a single locus is determined as likely harbouring a mutation. I will include a check and omit the phylogenetic tree inference in such a case.

I hope that helps!

Best, Jochen

[singerj]

Also note that the "-r" option was removed in the current version as intermediate results are not stored. Instead SCIPhI should be invoked several times with different seeds.

[tamaraprieto]

Dear Jochen,

Thank you very much for the help. Indeed the check would be very useful!

I was not expecting a reason like that one because my dataset is composed of 30 single-cells whole-genome at 5X. I detected 2000 mutations using SCcaller with default parameters and 7000 being less stringent. Do you think I could be parameterizing in a wrong way?

sciphi \
        -i ${RESDIR}/${REF}.fa \
        --in ${WORKDIR}/SciphiSamples.${PATIENT}.txt \
        -o Sciphi \
        -l 1 \
        --lz 0 \
        --ls 1 \
        --pr 0.0001 \
        --seed ${oddnumber} \
        ${WORKDIR}/Sciphi.${PATIENT}.mpileup

I have seen the -u is 0 by default.

Best,

Tamara

[singerj]

Dear Tamara,

The "-u" is not the problem, which is used to decide which mutations to include into the final VCF.

Could you paste the content of the "SciphiSamples.${PATIENT}.txt" file?

[tamaraprieto]

Dear Jochen,

Here you are

/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS/CLL04.Saliva.real.bam	BN
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.BM.T1.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.BM.T10.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.BM.T11.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.BM.T12.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.BM.T13.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.BM.T14.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.BM.T15.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.BM.T2.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.BM.T3.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.BM.T4.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.BM.T5.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.BM.T6.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.BM.T7.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.BM.T8.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.PB.T1.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.PB.T10.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.PB.T11.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.PB.T12.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.PB.T13.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.PB.T14.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.PB.T15.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.PB.T16.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.PB.T2.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.PB.T3.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.PB.T4.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.PB.T5.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.PB.T6.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.PB.T7.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.PB.T8.real.bam	CT
/mnt/lustre/scratch/home/uvi/be/tpf/CLL04/RESULTS//CLL04.PB.T9.real.bam	CT