Question for #archaeasky 🦠🧫 folks: I'm looking at a specific contig in an Archaeal genome with repeat elements that remain difficult to resolve even with 100~1000x coverage and minimum 10k read length. Anyone else dealt with something similar? What could I even be looking at? - ThreadSky

binomicalabs.org • 6 days ago

Question for #archaeasky 🦠🧫 folks: I'm looking at a specific contig in an Archaeal genome with repeat elements that remain difficult to resolve even with 100~1000x coverage and minimum 10k read length. Anyone else dealt with something similar? What could I even be looking at?

Comments

xabivc.bsky.social•5 days ago

Reduce the amount of input data by being more stringent with minQ and min length. It has helped me in some cases. Also, if using flye, there are a couple of options indicated for the 10.4+ that help

binomicalabs.org•5 days ago

I doubt it's the amount of input data. I'm working with subsampled reads sets of 100x, with minimum read length of 10k. Testing out 4~5 different assemblers and output shows the same variation- perfect assemblies except for that one contig that seems hypervariable.

xabivc.bsky.social•4 days ago

I'd still suggest playing with longer min length if possible and higher min phred score even at the cost of cov. Might be easier to correct afterwards

binomicalabs.org•4 days ago

True, I wonder if curating aligned reads and then making new assemblies from them with longer cutoff will change things. Will try, thank you!

binomicalabs.org•6 days ago

Also taking tips on best practices for cases like this!

surtlab.bsky.social•6 days ago

Just remember it might not actually be easy or possible to resolve given polymorphism in the population you’re trying to sequence (repeats and mobile elements do annoying things with genome structures and do so quickly)

binomicalabs.org•6 days ago

True true - in this particular case the difficulty is only with one contig. All the other genomics compartments (chromosome & plasmids) seem to assemble fine. I think polymorphism seemingly confined to one contig makes it worth the effort to dig in further

surtlab.bsky.social•6 days ago

(The way around this is sequencing from smaller population sizes and starting from pure strains…but then you get into philosophical discussions about what is the platonic ideal version of a genome)

christophstc.bsky.social•6 days ago

plato? hm...

benadler.bsky.social•6 days ago

Jill Banfields group is probably quite interested in this or knows what you're looking at. Happy to pass your contact info along if you DM me it.

binomicalabs.org•6 days ago

Thank you! Not sure if this would be important enough, but I can certainly use some insight.

Comments

Posting Rules

Reply