search

freeseek / mocha

MOsaic CHromosomal Alterations (MoChA) caller
MIT License
81 stars 23 forks source link

Written 0 variants for all contigs #34

Closed BJWiley233 closed 1 year ago

BJWiley233 commented 2 years ago

Hi,

Really scratching my head here with my mocha input. I have the phased annotated of my unphased file and I don't get any variants written to the allelic shift *as.bcf output. It worked in the WDL but I was trying to run on an input file I made and it no longer works. Using us.gcr.io/mccarroll-mocha/r_mocha:1.15.1-20220518 image. It's as if bcf_sr_next_line isn't able to read any lines. This is example of first line of input file along with a phased line:

chr1    629119  .       T       C       .       .       GC=0.4325;ALLELE_A=0;ALLELE_B=1 GT:BAF:LRR      0/0:0.0249828:0.0337044 0/0:0.00921902:-0.0882683
chr1    629990  .       G       A       .       .       GC=0.43;ALLELE_A=1;ALLELE_B=0   GT:BAF:LRR      0|0:0.975216:0.387888   0|0:1.03519:0.455783

And this is input:

batch=Batch_b001
bcftools +mocha \
      --genome GRCh38 \
      --input-stats /storage1/fs1/bolton/Active/projects/mocha/UKBB/arrays/$batch.ukb.sample.tsv \
      --output allelic_shift_mocha_vcfs/ukbb.$batch.as100.bcf \
      --output-type b \
      mocha_in/$batch.phased.100.min.bcf

and output

...
Using 61951 out of 61952 read variants from contig chr1
Written 0 variants for contig chr1
Using 60212 out of 60212 read variants from contig chr2
Written 0 variants for contig chr2
... (all contigs have "Written 0 variants for ..."
BJWiley233 commented 2 years ago

Well I was using the wrong docker 😄. Supposed to be using us.gcr.io/mccarroll-mocha/bcftools:1.15.1-20220518. I'm assuming you used apt-get to install bcftools in the r_mocha docker? If so we know something is wrong with the htslib library and synced_bcf_reader.c isn't working there...