Dear community members,
We have available WGBS data for a species. We used Bismark to map the reads to the reference genome. Unmethylated lambda phage DNA was spiked-in, to estimate the bisulfite conversion error rate. My question is, how can I use the information of this error rate to distinguish truly methylated reads from false positives?
I have seen that many people use a binomial distribution to achieve this, and apply a Benjamin-Hochberg correction citing their 1995 paper (https://www.jstor.org/stable/2346101?seq=1#page_scan_tab_contents).
Is there a tool or an easy way to do this? Excuse my ignorance but I am really new to the field.
Thank you for your time,
Panos
You can try to estimate your conversion efficiency of each fragment. The conversion efficiency is the number of C present on your sequence and not aligned to a CG on the genome of reference.
Dear Panagiotis,
i think you are confusing two concepts.
The first concept you are talking about is the bisulphite conversion rate that can be inferred as you have proposed or simply looking at the C to T conversion in the mitochondria genome. Since mitochondria is almost all unmethylated you will expect at least 99% of its genome converted from C to T.
The second concept is the differentially methylation analysis that allows you to distinguish which cytosines are highly or lowly methylated in between two or multiple conditions. Once you have detected differentially methylated regions (or DMRs https://en.wikipedia.org/wiki/Differentially_methylated_regions) you have to perform multiple testing in order to rank the p.value of each DMR discarding possible false positive results.
I hope this can help you.
Best
- Badges
- Science topic
- Similar topics
- Bioinformatics
- Bioinformatics and Computational Biology