One HVR is a genus-wide feature and anchored by tRNA genes. One HVR is conserved across the genus at a fixed chromosomal position. The internal GC gradient is a “bathtub curve” and is evidence for a two-ended insertion model where phages repeatedly deposit surface modification genes at the tRNAs.
oops forgot to tag Torben Nielsen @tnn1.bsky.social !
The preprint is long ( >10,000 words!), but it’s the paper that we wanted to write and we couldn’t figure out a way of making it shorter without losing major points. And yes, there’s more to come.
The Pelagibacter pangenome is likely driven by two distinct mechanisms: phage-mediated surface coat diversification concentrated in the HVR, and independent genomic island insertions carrying metabolic, defense, and transport genes throughout the rest of the chromosome.
Synteny is low across the genus. Within genome pairs have high synteny but between species synteny rapidly decreases with ANI.
With a 135 complete, circular Pelagibacter genomes, we have answered some of the most outstanding questions about what is often considered the most abundant organism on the planet, with roughly 10 million times more individuals in the ocean than stars in the universe. Check out our preprint.
How much Pelagibacter diversity were we missing? Quite a bit. We clustered the genomes into 52 species-level groups using a 95% ANI threshold; 44 (85%) of these are novel.
Why are Pelagibacter genomes so difficult to assemble?
You need a high amount of long read sequencing, plus the right assembler (thanks @jimshaw.bsky.social, @lh3lh3.bsky.social, and @mgmarin.bsky.social) to disentangle a universally conserved hyper-variable region and highly conserved genes.
Auxotrophies appear to phylogenetically structured, not random.
Auxotrophies cluster on the tree, rather than being randomly scattered, suggesting that these are not neutral losses and possibly ecological specialization.
We recovered 106 genomes from the San Francisco Estuary and 29 we collected from public databases. Big thanks to Miten Jain, USGS, and @jgi.doe.gov in @biosci.lbl.gov.
