hippo-yf
commented
3 months ago 1, using the sorted-deduplicated bam files as input to generate the ATCGmap files is ok 2, use ATCGmap files in cgmaptools/bsgenova for variant calling 3, "My surprise was that the vcf file had degenerated bases.", you mean base wildcards in the vcf file generated by cgmaptools ? 4, if you are interested in differentially methylated sites/regions, variant calling is optional 5, according to my experience, model organism or not do not affact too much to choose DMR software, 6, https://pubmed.ncbi.nlm.nih.gov/36147684/ for an evaluation of mapper
desmodus1984
Hi,
I am analyzing methylation data- finding differentially methylated sites/regions, and I wanted to inquire about suggestions whether my input is good or if I should preprocess with better software. My collaborator enzyme-converted DNA with the NEBNext® Enzymatic Methyl-seq Kit](https://www.neb.com/en-us/products/e7120-nebnext-enzymatic-methyl-seq-kit, and sequenced with Nova-seq. I removed adapters with trimgalore!, and then mapped the reads with bwa-meth, bam converted and sorted, and then used Picard to eliminate duplicate reads.
According to two papers, the bayesian model of cgmaptools is the suggested tool for non-model organisms. Hence I used the sorted-duplicate-removed bam files and generated the ATCGmap files and did the variant calling. My surprise was that the vcf file had degenerated bases.
Thus, I wanted first to ask if the sorted-deduplicated bam files could be used as input or if I should use the ATCGmap generated from them.
Thanks;