Segmentation fault in population calling

[kendaida]

Hi, I am doing population calling using 100 samples as mentioned here (#77 ). It runs good but when it comes to force calling, I had an error like below. I assume the size of the VCF using in "Ivcf" may cause the error. It is about 32G. Would you please teach me how to resolve this error?

Thanks, Ken

CODE

cuteSV ${BAMFILE} ${REF} ${OUTPUT} ${WORKDIR} -Ivcf ${FORCECALL_VCF} \
             --sample ${SAMPLENAME} --max_size -1 --max_cluster_bias_INS 100 --diff_ratio_merging_INS 0.3 \
             --max_cluster_bias_DEL 100 --diff_ratio_merging_DEL 0.3 --min_read_len 50

ERROR

2022-09-15 11:18:50,022 [INFO] Finished chr17:40000000-50000000.
2022-09-15 11:18:50,122 [INFO] Rebuilding signatures of structural variants.
2022-09-15 11:18:51,163 [INFO] Check the parameter -Ivcf: OK.
2022-09-15 11:18:51,163 [INFO] Enable to perform force calling.
CuteSV_forcecall.sh: line 10: 15614 Segmentation fault      cuteSV ${BAMFILE} ${REF} ${OUTPUT} ${WORKDIR} -Ivcf ${FORCECALL_VCF}  --sample ${SAMPLENAME} --max_size -1 --max_cluster_bias_INS 100 --diff_ratio_merging_INS 0.3 --max_cluster_bias_DEL 100 --diff_ratio_merging_DEL 0.3 --min_read_len 50

[Meltpinkg]

Hello, @kendaida

In our experiments we haven't met the errors like segmentation fault. I think it may due to the large size of Ivcf (since I haven't used such a large size). You can split the Ivcf file into various parts (like divide by the chromosomes) and check whether it can run successfully in each part. And finally, paste the various parts together.

By the way, can I know how many SVs and how many samples are contained in the 32G Ivcf? And which version did you use in the population calling?

Best regards, Shuqi

[kendaida]

Thank you Shuqi, I got it. I used just one hundred samples but did use "--max_size -1" options. The version is the lateset v2.0.1.

Best, Ken