Per-amplicon coverage plot about viralrecon HOT 5 CLOSED

I think samtools has been writing code to do the same primer trimming and get stats for each amplicon - samtools/samtools#1261. Not quite sure of other tools though!

from viralrecon.

We made something similar, not as cool but maybe you can find it helpful. We have the scripts here https://github.com/BU-ISCIII/SARS-Cov2_analysis/tree/master/Illumina-amplicons/13-amplicon_plots
We perform first bedtools coverage to obtain the depth and coverage of the amplicons, for this we use a modified .bed with the start and end position of the amplicons for both forward and reverse primers. In the case that for the same amplicon there were both normal and altered positions, we took the longest one. To create this file we just took the original .bed file and parsed it, so for the same amplicon instead of having one line for the RIGHT and another line for the LEFT we have just one amplicon per line. Then we use R to create the plots.

from viralrecon.

The log only reports reads from which primers were trimmed off but with NextEra to get the true counts you will have to check reads that start and end within the left and right primers of an amplicon. Will require some tweaking of code to log that as well but unfortunately I'm rather occupied with other SARS-CoV-2 things at the moment.

from viralrecon.

No worries! Thanks alot for the rapid response as always @gkarthik 😎 Agreed, crazy times! Are you aware of any scripts I can use to do such a thing or get an idea as to which tools to use to generate the plots? I have an idea but would rather avoid re-inventing the wheel.

from viralrecon.

Initial implementation in #119

So I wrote a custom script to collapse the amplicon bed files based on some sort of standard naming convention for left and right primers in the BED file. Works with all of the ARTIC BED files.

Was very easy to use mosdepth to get genome-wide coverage in 200bp windows as well as per-amplicon coverage and then to plot in R. Some examples in this comment.

from viralrecon.