The abstract is read at the publisher; the summary is JClub's.
zsasa reproduces FreeSASA atom-level solvent accessible surface area on 4,370 Escherichia coli AlphaFold models while matching results and running much faster on proteome-scale inputs.
- Why it matters: Existing SASA tools lag behind current structural volumes and formats, limiting speed and scalability for proteome-scale structure and trajectory analyses.
- What they did: The open-source zsasa engine implements established SASA algorithms with multicore and SIMD acceleration, then benchmarks it against FreeSASA, other tools, and MDTraj across single structures, human AlphaFold proteomes, and trajectories.
- The result: On 23,586 human AlphaFold proteome structures, zsasa is 3.3-fold faster than FreeSASA in parallel using a third or less peak memory, and in approximate mode is 10-fold faster with average total SASA within 0.7%; on trajectories it is 12- to 52-fold faster than MDTraj.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.