compatibility with HiFi reads

[dabitz]

Hi Pierre,

Does discosnp works with PacBio HiFi (>99% accurate) reads?

Best, André

[pierrepeterlongo]

Hi André.

We have no feedback of using such sequencing data with discoSnp. However, the mapping step (reads on predicted variant sequences) authorizes only substitutions. Hence I expect bias in term of quantification.

Have you tried this?

[dabitz]

Not yet. Would it be possible to set up the mapping strategy with minimap within discosnp script? If the coverage is high enough and the mapping step is stringent enough I would expect discosnp to work, right?

[pierrepeterlongo]

Would it be possible to set up the mapping strategy with minimap within discosnp script?

No I don't think so. Mainly for two reasons:

• [bad reason] the analyse of the mapping step for creating the vcf depends on the output format of the mapping step.
• [good reason] the mapping step must be able to map read with errors but with 100% precision on the variant position(s) as predicted by discoSnp. Else, each mapped read as high chances to map ref and alt sequences leading to bad estimations of allele frequencies.

However this would be highly interesting to make a try to parse minimap outputs to retrieve this precision between each pair of sequences.

Any input is warmly welcome