Hi all,
I run dindel as part of workflow which I do for dna seq, which large percent is exome samples. I've noticed something quite odd with regards to dindel stage 3.
I run stage 3 parallel using cluster. I split the input with '--numWindowsPerFile 1500' currently. Most of the executions take between 20 and 50 minutes to run. Some take more. However I've noticed that in many cases there is one that takes a lot more time (can take several days). That seems to happen in situations where the chromosomal position is in chr 10 and 42Mb. Has anyone else noticed anything similar? There seems to be high number of haplotypes printed in log*, but I did check this for each file in one sample and mean haplotype of all lines per file doesn't seem to follow execution time that file.
* eg the it has something like this:
[empiricalDistributionMethod] Number of haplotypes: 136
I run dindel as part of workflow which I do for dna seq, which large percent is exome samples. I've noticed something quite odd with regards to dindel stage 3.
I run stage 3 parallel using cluster. I split the input with '--numWindowsPerFile 1500' currently. Most of the executions take between 20 and 50 minutes to run. Some take more. However I've noticed that in many cases there is one that takes a lot more time (can take several days). That seems to happen in situations where the chromosomal position is in chr 10 and 42Mb. Has anyone else noticed anything similar? There seems to be high number of haplotypes printed in log*, but I did check this for each file in one sample and mean haplotype of all lines per file doesn't seem to follow execution time that file.
* eg the it has something like this:
[empiricalDistributionMethod] Number of haplotypes: 136