|
Post by brookb on Nov 12, 2019 3:24:38 GMT
Hello, I am trying to use TRUFFLE to find pairwise IBD segments. The software fails to finish with the --pair option (using sample IDs from multi-sample VCF input to specify pairs). The program completes successfully on the same input VCF when the --pair option is not specified, here is the output of the failed command:
[*] Finished reading input files: 1449645.07 ms
- Expected IBS for unrelated pairs: p.IBS0 = 0.0660 p.IBS1 = 0.4274 p.IBS2 = 0.5067
- Likelihood based region estimate: 1340 p=7.99196e-11
(ibs_est) - 549 p=0.934037 perr=0.000000
(ibs_est) - 56 p=0.506675 perr=0.000000
(ibs_est) Thresholds for defining regions as IBD1 = 549, IBD2 = 56
(ibs_est) Inflation at edges IBS1 = 28.3, IBS2 = 1.9
- Genome size: 14545555.294000
- gs_threshold=41057.644051
- (ibs_est) Adjust segment size to 41057 markers (from 549)
- IBS1 threshold : 0.08 % of genome (41057 markers)
- IBS2 threshold : 0.01 % of genome (6842 markers)
[*] Starting IBD segment loop
- Allocation genotype vector transposed npeople=18 nvars=54564296
- GenotypeMatrixTransposed: allocating 936 MB of memory
[*] Genotype pre-processing duration: 1504475.52 ms
- Compute IBD by IBS: (cpu=1/1) Nind = 17 Nvar = 54564296
Running pair -1 -1
Segmentation fault
Any idea why this is failing, or how to fix? Thanks, Brook
|
|
|
Post by apostolis on Nov 28, 2019 15:06:17 GMT
Hi, the pair was not found in the input VCF file, can you double check that the sample names match? We will look to report the error in future versions.
|
|