Hello,
I am using vds combiner:
gvcfs =
with hl.hadoop_open(path_to_input_list, ‘r’) as f:
for line in f:
print(line)
gvcfs.append(line.strip())
combiner = hl.vds.new_combiner(
output_path=‘/opt/buzkova/final/datasetFinal2.vds’,
temp_path=‘/opt/buzkova/final/my-temp-bucketFinal2’,
gvcf_paths=gvcfs,
use_genome_default_intervals=True,
reference_genome=‘GRCh38’
)
combiner.run()
On haploid alleles (sites) in chrX and haploid alleles (sites) in chrY Hail combiner do not pass LPL entries instead there are stored empty arrays as you can see on the following picture.
This cause problem with command split_multi = hl.vds.split_multi(vds) and the following error shows up when trying split haploid alleles
Error summary: HailException: array index out of bounds: index=0, length=0
… line 189, in \n .map(lambda idx: old_entry.LPL[idx]
I do not want to drop LPL because I need to recalculate the GQ values based on PL.
My raw gVCFs on haploid sites has PL field. I am sending as an attachment a text file (gvcf_examples.txt) with part of chrX at position 2781514-2782567 (haploid entries) for all 10 files that I am using in analysis. And I also attached file (gVCF_head_and_contig_info_example_gVCF_1.txt) that has the head and contig info of the first gVCF - GVCF_1_chrM.gvcf.bgz
gvcf_exapmles.txt (30.4 KB)
gVCF_head_and_contig_info_example_gVCF_1.txt (172.9 KB)
I will be grateful for your help,
Veronika.