Back

Mapping the Metalloproteome of Deinococcus indicus DR1 through Integrative Structure and Function Annotation

Ramesh, S. D.; Vasan, G.; Senthilkumar, S.; Thambiraja, M.; Sethi, D.; Priyadarshini, R.; Yennamalli, R. M.

2025-07-17 bioinformatics
10.1101/2025.07.11.664487 bioRxiv
Show abstract

Deinococcus indicus DR1 is a rod-shaped bacterium isolated from the Dadri wetlands (Uttar Pradesh, India) that tolerates ionizing radiation and arsenic. The molecular basis of its wider heavy-metal resilience, particularly among the 1017 out of 4128 proteins still annotated as hypothetical, remains unclear. We performed a proteome-wide structural and functional survey to address this gap. All the 4128 proteins were modeled with AlphaFold2, yielding very-high-confidence structures (pLDDT [≥] 90) for 2145 sequences. CATH/InterPro analysis assigned domains to 2735 proteins. Functions were predicted by combining DeepFRI (graph neural-network), MorphologFinder (Foldseek plus EggNOG-Mapper), and existing GenBank annotations. The integrated workflow suggests that more than 100 previously uncharacterized proteins may bind or transport arsenic, chromium, cobalt, copper, iron, manganese, molybdenum, nickel, or zinc, indicating a metal-handling capacity that extends beyond the known ars operon. Recurrent domain architectures that include P-loop NTPases, Rossmann folds, GNAT acetyl-transferases, and sensor modules (CHASE, PAS, GAF) point to coordinated redox regulation and efflux pathways. Twenty high-confidence metal-binding candidates have been prioritized for experimental validation through expression, mutagenesis, and knockout studies. All structural models, domain assignments, and query tools are available at https://deinococcus.in, providing a resource for future investigations of heavy-metal tolerance in this organism.

Matching journals

The top 12 journals account for 50% of the predicted probability mass.

50% of probability mass above

"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.