forterre

pforterre.bsky.social

16 September 2025 03:11

The phylogeny in figure S1 shows clearly that the rare rubulins in Asgard and Bacteria have been obtained from proto-eukaryotes. In my last mlife paper, I document extensive LGT between Asgard and proto-euka, even for univerdal proteins.

Steven Robbins

stevenjrobbins.bsky.social

05 September 2025 00:59

This is a cool study, just wanted to highlight that the problem its solving isn't low abundance directly. Say your microbe of interest has real abundance of 0.5% & 29% GC. Our analysis shows that the abundance in Illumina reads might be as low as 0.025% due to bias. www.biorxiv.org/content/10.1...

Steven Robbins

stevenjrobbins.bsky.social

05 September 2025 00:59

Side note: One often thinks they're not recovering low-abundance taxa because they're low-abundance. But with Illumina, it's also likely because they're low-GC and Illumina has strong GC bias. You can fix this with wetlab intervention like this, or using ONT instead, which doesn't have a GC-bias.🧵

Steven Robbins

stevenjrobbins.bsky.social

05 September 2025 00:59

Does ONT completely solve all metagenomics problems, no. But not having a GC bias helps a lot. At least in our hands, we recovered a lot more low-GC taxa with ONT than Illumina. And what's "most abundant" in our Illumina vs ONT reads aren't even similar...the differences always corresponding to GC.

Steven Robbins

stevenjrobbins.bsky.social

05 September 2025 00:59

GC-bias also hugely distorts one's view of what is "abundant" in the first place. Nothing with low-GC will almost ever come up as abundant in Illumina because of intrinsic GC bias. Unless you did microscopy counts first, you never knew what was truly abundant.