Hi,
I am very new to long read sequencing and is confused about which are the good variants callers (SNVs and INDELs) available for data from nanopore sequencing.
Currently I have a fastq file with reads from two genes.
I came across some tools like Medaka and NanoCaller. But I wish to know if there are any other better options.
Please give some suggestions.
Thank you.
2 answers
ONT offers a human variation nextflow pipeline: https://github.com/epi2me-labs/wf-human-variation They are using clair3 and sniffles.
Clair3 and longshot are both useful for nanopore long read SNP calling. As you point out, Nanocaller is also a decent option.
Keep in mind SNPs will be far better from Q20 or duplex reads. If you have high coverage (is this amplicon data?) consider filtering out poor reads before alignment, eg keeping those with a Q score of 10 or more (if Q20 data).
Sniffles2 is a SV caller so not applicable for SNPs.
Finally - I would stick to SNPs, and not indels, since ONT, like most seq techs, performs poorly on indels.
Log in to answer this question.