sttongjai commented 1 year ago

Hello,

I was using MethylDackel via Conda to extract CpG from a BAM file created by bwa-meth. After my first attempt of extraction CpG which was successful, my second attempt, aiming at repeating the previous extraction, was not.

I received an error message telling that the program could neither load an index file of a reference genome (hg38 with no alternative chromosomes) nor build one for it. It also suggested that the reference genome file could be corrupted. However, the index file from the previous run was in the same directory as the reference genome.

Could you please let me know what's a solution to this problem.

Thank you very much,

Siripong

P.S. Please see a list of programs and their versions installed in my conda environment below.

akshaydinesh26 commented 3 months ago

I am also facting the same issue. When the program was run separately for each file it was working but not using bash script

dpryan79 commented 22 hours ago

One "fun" issue is that if there is no index for the genome (a .fai file) then running this or any similar tool in parallel will result in each process recreating that file at the same time and overwriting each-others results. My guess is that something like that happened in both of these cases. There's really nothing that can be done within MethylDackel itself to guard against that, "just" ensure that the fasta file is indexed with samtools faidx before running.

dpryan79 / MethylDackel

Could not repeat a CpG extraction with the same reference file and its index #148

Name Version Build Channel