Saphyr support?

[RenzoTale88]

Good morning,
I was wondering if the software support Saphyr optical maps and DLE-1 reads?
Thank you in advance
Andrea

[StefanoLonardi]

Ciao Andrea, OMGS should work with DLE1.
You will had to add the DLE1 restriction enzyme in the fa2cmap_multi.pl perl script as below.

my %enzyme = (
	"BspQI" => "GCTCTTC",
	"BbvCI" => "CCTCAGC",
	"BsmI"  => "GAATGC",
	"BsrDI" => "GCAATG",
	"bseCI" => "ATCGAT",
	"DLE1" => "CTTAAG"
);

If you have issues with OMGS feel free to contact directly Weihua at panweihua@caas.cn.
Stefano

[RenzoTale88]

Ciao Stefano,
Thank you for this, I've just tried and it seems to be working just now. I'll come back again in case I come across any issue.
A quick question: in my case, I have haplotype resolved contigs generated using hifiasm on HiFi reads and HiC sequencing. Do you think the OMGS approach will be able to generate scaffolds even in this scenario?
Andrea

[StefanoLonardi]

Ciao Renzo,
in my experience HiFi+bionano map is superior to HiFi+HiC, but I do not have much experience on HiFi+HiC+Bionano. The optical map is more than just for scaffolding. It can be very helpful in detecting chimeric contigs (i.e., mis-assemblies) and estimating how much of the genome you are missing in your sequence assembly.

By the way the bionano hybridscaffolding pipeline is also very good. It is not hard to use using docker.

Stefano

[RenzoTale88]

@StefanoLonardi thanks for this. Inmy specific case, we used the HiC not to scaffold, but to generate a haplotype-resolved assembly. In effect, we get two separate genomes, of similar lengths, representative of the two distinct haplotypes (genome1.hap1.fa and genome1.hap2.fa). However, these two sets of contigs are not necessarily split by parent (i.e. hap1 is not always the e.g. maternal), which is why I wanted to use the OM to try scaffold them.
I've been trying the bionano HybridScaffolding pipeline, but we end up with weird chimeras, perhaps due to the dual representation of the haplotypes.

[StefanoLonardi]

Ciao Renzo, we never tested OMGS on haplotype-resolved assemblies, so this is uncharted territory.

It might be a problem that requires a new tool. 😄