Litcius/Paper detail

GPU-accelerated homology search with MMseqs2

Felix Kallenborn, Alejandro Chacón, Christian Hundt, Hassan Sirelkhatim, Kieran Didi, Sooyoung Cha, Christian Dallago, Milot Mirdita, Bertil Schmidt, Martin Steinegger

2025Nature Methods37 citationsDOIOpen Access PDF

Abstract

Rapidly growing protein databases demand faster sensitive search tools. Here the graphics processing unit (GPU)-accelerated MMseqs2 delivers 6× faster single-protein searches than CPU methods on 2 × 64 cores, speeds previously requiring large protein batches. For larger query batches, it is the most cost-effective solution, outperforming the fastest alternative method by 2.4-fold with eight GPUs. It accelerates protein structure prediction with ColabFold 31.8× over the standard AlphaFold2 pipeline and protein structure search with Foldseek by 4-27×. MMseqs2-GPU is available under an open-source license at https://mmseqs.com/ .

Topics & Concepts

Computer sciencePipeline (software)MIT LicenseGraphics processing unitComputational biologyGraphicsSequence homologyHomology (biology)Protein structureNearest neighbor searchSearch algorithmSearch engineDatabase search engineSequence databaseData miningHomology modelingProtein structure databaseSequence alignmentComputer graphicsProtein methodsParallel computingComputer graphics (images)AlgorithmGenomics and Phylogenetic StudiesMachine Learning in BioinformaticsGlycosylation and Glycoproteins Research
GPU-accelerated homology search with MMseqs2 | Litcius