[DISCUSSION] Methylation processing in anvio

[Ge0rges]

Hello MerenLab,

While at EBAME I was encouraged to open this discussion by @FlorianTrigodet and @ivagljiva, and bring it to @semiller10's attention.

DNA methylation is on the verge of constituting a significant amount of sequencing data. Most methylation detection assays till now have relied on bisulfite sequencing. However this method is limited to certain types of methylation, and is not useful in the context of understanding community level DNA methylation. While PacBio offers very accurate methylation data, it is expensive. Nanopore promises to make epigenetic studies widely accessible thanks to it recent embrace of DNA modification detection. They already support 2/3 types of DNA methylation, and expect to support all 3 by next May. Moreover, Nanopore's method enable for methylation detection in any nucleotide context (sequence).

Thus I was wondering if there is any interest in adding DNA methylation support to anvio. This would consist at least in reading in methylation data from BAM files and match it to the corresponding read. More convenience functions could include things like:

• Getting a "methylated MAG" by mapping reads back onto the MAG into anvio (even if mapping is done externally)
• Getting differentially methylated regions (DMR) of a MAG (see MethylKit and modkit).
• Correlating DMR patterns between samples
• Getting the annotation of a DMR

These could then be very nicely visualized using Anvio which would be a big benefit versus existing software.

I would be very happy to work with the Anvio team towards making this happen, though my knowledge of the codebase is currently limited.

[semiller10]

Hi @Ge0rges, thanks for the pertinent suggestion. I'm also aware of work on RNA modification profiling by Nanopore. Anvi'o is currently limited in its ability to track genomic features beyond gene calls, but we are planning to soon implement a general framework for storing the coordinates of any type of genomic feature, like introns and exons, promoters, or modified nucleotides. Once we implement this, we'll get back in contact with you regarding what to do with modifications -- all of your suggestions sound perfect, and I'm glad to hear that there's a gap in the field that anvi'o can fill.

[Ge0rges]

Sounds great! RNA modifications are definitely becoming a priority for ONT, and I think will open up a lot of interesting analyses. Let me know if I can be of any help.