mbias result is different between bismark and bwameth output

[nnlrl]

Hello,
I am doing some wgbs analysis. I find that the mbias results obtained by using different bisulfite reads aligners(bwa-meth and bismark) for the same sample are completely different. This is my command. Can you tell me what is wrong

# bismark
bismark  -X 700  --multicore 8 --genome_folder /path/to/genome/ --output_dir  /path/to/output/ --nucleotide_coverage --bowtie2   -1 /path/to/read1.fq.gz -2 /path/to/read2.fq.gz
# bwa-meth
bwameth.py  --reference /path/to/genome/genome.fa -t 12 
/path/to/read1.fq.gz /path/to/read2.fq.gz

This is BAM output using Bismark and Methyldackel mbias

This is BAM output using bwa-meth and Methyldackel mbias

[y9c]

Hi @nnlrl, this is quit interesting. Is it because bwa-meth filter partially converted reads? Could you share the conclusion you draw from this comparison?