Utilizing an artificial intelligence system to build the digital structural proteome of reef-building corals

Abstract

Background Reef-building corals play an important role in the marine ecosystem, and analyzing their proteomes from a structural perspective will exert positive effects on exploring their biology. Here we integrated mass spectrometry with newly published ColabFold to obtain digital structural proteomes of dominant reef-building corals. Results Of the 8,382 homologous proteins in Acropora muricata, Montipora foliosa, and Pocillopora verrucosa identified, 8,166 received predicted structures after about 4,060 GPU hours of computation. The resulting dataset covers 83.6% of residues with a confident prediction, while 25.9% have very high confidence. Conclusions Our work provides insight-worthy predictions for coral research, confirms the reliability of ColabFold in practice, and is expected to be a reference case in the impending high-throughput era of structural proteomics.

MineProt: a stand-alone server for structural proteome curation

“…The current version of MineProt Toolkit can directly transform the raw outputs of AlphaFold (up to v2.3.x) and ColabFold (up to v1.5.x) into a fully functional website, while supporting any other data compatible with its standard as well. The CP-8382 dataset, containing 8166 protein structures with computing cost of 4060 GPU hours ( 5 ), needs only 92 min for full-flow curation (tested with 10 threads of Intel Core i5-10 400 on a WDC WD20EZAZ-00 L).…”

“…When AI systems are rapidly generating structure predictions, problems arise about how to curate these novel high-throughput data. It is difficult for public databases like AlphaFold DB to curate all new data without delays, while it might be hard for most researchers to develop a similar website from scratch to publish their data in a more intuitive way ( 5 ). In the history of bioinformatics, a similar situation occurred when next-generation sequencing technology emerged, and it is a stand-alone server called SequenceServer ( 6 ) that provides an excellent solution for this issue.…”