search

tgen / lumosVar2

Calls somatic SNVs, indels, and allelic copy number jointly across multiple samples from the same patient. These can be standard tumor/normal pair, longitudinal samples, primary/met, etc. Can also be used for tumor only calling, ideally with a high tumor content and a low tumor content sample.
MIT License
10 stars 1 forks source link

Error using parfor in segmentData when running two chromosomes #5

Closed a-jartseva closed 4 years ago

a-jartseva commented 4 years ago

Hello, I am doing a test run with just two chromosomes, 21 and X, as described in the instructions (see https://github.com/tgen/lumosVar2/issues/2#issuecomment-638215094 for why I am running the X chromosome, too). It seems like there is an issue using the parfor command:

Error using parfor_range_check (line 28) 
The range of a parfor statement must be increasing consecutive integers.  See Parallel Computing Toolbox, "parfor".

Error in segmentData (line 37)

Error in lumosVarMain (line 178)

As far as I understand, in this line https://github.com/tgen/lumosVar2/blob/81b31d7230dfd4ede1e2e015252d717f726c91d6/src/segmentData.m#L36 a list of chromosomes to run the loop on is created; however, if not all chromosomes are analysed, the list is not made of consecutive integers, which is not compatible with parfor. Could you fix this bug?

Thanks a lot, Alex

rhalperin commented 4 years ago

Thanks for brining this issue to my attention. LumosVar is really intended for analysis on a full exome. While I do suggest running on only one chr in the instructions as a quick installation test, I don't see a practical use case for running one only two chr.

a-jartseva commented 4 years ago

Yes, that makes sense, but it is nice to be able to do a quick test on one or two chromosomes rather than spending a long time analysing the whole exome and realising that something is not working down the line. I suppose if the issue in the other thread (https://github.com/tgen/lumosVar2/issues/2#issuecomment-638215094) is resolved, then I'll be able to just run one chromosome and not run into this problem. Thank you!