I'm very excited to share our manuscript for Cenote-Taker 3. We show that (1) CT3 quickly and accurately annotates previously uncatalogued virus genomes and (2) finds divergent virus genomes in contiguous assemblies. Virome benchmarks are SO hard, but I think we came up with clever and useful ones.

Hey, if you're interested in how my team and I sequence all the human and animal viruses in community wastewater, come register for the webinar I'm delivering next week! 🦠🖥️🧬

Finally, while the effect was not huge, we found that both bacterial and phage community change was slower in participants who eventually developed type 1 diabetes in a nested case-control study. This may say something out the interaction of the microbiome and the immune system. 6/6

We also see that distinct phage species (SGBs) follow different temporal patterns, rising and falling across the first years of life in our cohorts. Using the phage data also helps random forest classifiers to understand our data. 5/

Applying this method to over 12,000 gut metagenome sequencing datasets from the TEDDY study, which follows kids prone to type 1 diabetes, we found that phages are more diverse than their host bacteria and phage populations change faster than bacteria. 4/

Phages are hard to measure by sequencing. Inspecting alignments of metagenomic reads to phage genomes, we often see that a small region has high coverage, indicating a spurious detection. So we decomposed thousands of phage genomes into marker genes (capsids, terminases, etc) and aligned to those.2/