search

KolmogorovLab / hapdup

Pipeline to convert a haploid assembly into diploid
Other
85 stars 8 forks source link

Support for CLR reads? #32

Closed ptrebert closed 1 year ago

ptrebert commented 1 year ago

Hi, I was a bit surprised to see that there is no support for PacBio CLR reads (out-of-the-box at least). I have a bunch of flye CLR assemblies; do you think it's worth a shot to run hapdup on those (if so, using which setting)? Best, Peter

ptrebert commented 1 year ago

Just to be more precise, the readme claims compatibility with CLR ...

[...] HapDup works with either Oxford Nanopore or PacBio (HiFi or CLR).

... but the presets only apply to ONT and Hifi (as well as the hard-coded default PEPPER/MARGIN model files). This is what confuses me.

mikolmogorov commented 1 year ago

@ptrebert Sorry for my late response!

It would be difficult to add PacBio CLR support, as neither PEPPER nor Margin have modes for this type of data. With most groups quickly switching to either ONT or PacBio HiFi, there is little incentive to develop specialized options for Pacbio CLR, unfortunately. Have you tried the HiFi / ONT modes, do they work at al?

Misha

mikolmogorov commented 1 year ago

The manual citing Pacbio CLR is indeed my mistake - it should be removed..

ptrebert commented 1 year ago

I fully agree on the fact that the field has largely abandoned CLR reads, I just happen to have a request on my desk about analyzing some old-ish datasets. Yes, I did try to run hapdup with the existing model, but it (predictably) failed.

So, overall, I'll close this ticket now; except for the minor correction to the readme, there is not much more discuss I guess. Thanks.