Leak CURBER enzyme classification: homology transfer

Each query sequence is aligned against all unique training sequences with DIAMOND (--very-sensitive). Hits whose bitscore is at least 80% of the best hit vote for their EC labels, with weight bitscore^4 split evenly across a hit's labels. The EC with the largest vote is predicted. Queries with no hit get the most frequent training EC.

train.py --train train.parquet builds weights/train.dmnd, weights/train_labels.parquet and weights/fallback.json.
predict.py --input INPUT --output OUTPUT runs inference on CPU.
bin/diamond is DIAMOND v2.1.10 (linux x86-64 static binary, https://github.com/bbuchfink/diamond).

EC vocabulary: ExPASy ENZYME release 02-Sep-2026, SIB, CC BY 4.0.

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support