Aaaaand this one is from Curtis!

10/ Special thanks to our collaborators @hillspet.bsky.social for their contributions and feedback throughout the development of Parathaa.

9/ Any feedback or issues can be directed to our forum at: forum.biobakery.org.

8/ Parathaa is openly available for use by the community and can be used to assign amplicons to both SILVA and GTDB taxonomy. github.com/biobakery/pa...

7/ Moreover, application of Parathaa to real-world data identified cases where taxonomic labels found within databases were likely incorrect or incompatible with sequence phylogeny.

6/ We found that Parathaa performed in F1, precision, and recall just as well or better than commonly used amplicon assignment methods while also allowing for finer species resolution.

5/ We validated Parathaa on seven synthetic Silva datasets and historic holdout data from GTDB using both full‑length 16S rRNA amplicons and amplicons of the V1V2 and V4V5 regions.

4/ Parathaa works in two main steps:
- Build an optimized, primer‑specific map linking phylogenetic placement to taxonomy.
- Apply this map to assign zero or more taxonomic labels to each sequence.

3/ Parathaa aims to address these issues by directly modelling taxonomic sequence ambiguity within sequenced amplicon regions: i.e. allowing for assignments to multiple taxonomic labels when phylogenetically warranted.

2/ Amplicon data often contains unique biases due to primer incompatibilities and the inability to differentiate between certain microbes due to low sequence variability, especially at the species level.

Eric also teaches graduate courses in Bioinformatics and Python programming at the School alongside the Lab’s own microbiome training workshops. Eric is an avid tabletop gamer outside of work (and inside, where appropriate) and never misses an episode of Saturday Night Live :)