Zero-shot ecological annotation of microbial genomes with myLLannotator accelerates scientific discovery
Lee, A. L.; Sharma, A.; Maddamsetti, R.
Show abstract
Large language models (LLMs) are promising scientific assistants, but many models remain inaccessible for resource-constrained scientists. To address this, we present myLLannotator, a python package for metadata annotation built on llama3.2-3B. myLLannotator labels 18,000+ genomes in ~2 hours on a laptop, reproducing a discovery involving substantial manual annotation. We use myLLannotator to discover that duplicated genes are depleted in endosymbiotic bacteria, demonstrating its power to accelerate hypothesis testing and discovery.
Matching journals
The top 7 journals account for 50% of the predicted probability mass.
Similar papers in this journal
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.