Version 0.15 / bp.p_ctg.gfa versus bp.hap?.p_ctg.gfa N50

[chklopp]

Hi,

For several assemblies we found better N50 metrics for bp.p_ctg.gfa than both bp.hap?.p_ctg.gfa.
Why are why losing continuity when separating haplotypes?
Is it pertinent to scaffold bp.hap?.p_ctg.gfa with bp.p_ctg.gfa?

[chhylp123]

bp.hap?.p_ctg.gfa dose not join different haplotigs while bp.p_ctg.gfa joins some of them. Therefore bp.p_ctg.gfa usually has higher continuity at the expense of much fragmented a_ctg.gfa. We think fragmented a_ctg.gfa is not such useful so that half of information might be lost. Another difference is that bp.p_ctg.gfa tends to keep sex chromosomes at once. Just curious: how do you scaffold bp.hap?.p_ctg.gfa with bp.p_ctg.gfa?

By the way, version 0.15 has some minor bugs so that it's better to re-output assemblies with v0.15.1 or current github HEAD.

[chklopp]

The easiest way is to align bp.hap?.p_ctg.gfa.fa on bp.p_ctg.gfa.fa with dgenies (http://dgenies.toulouse.inra.fr) and export the query organized as the reference.