Back

AlphaFast: High-throughput AlphaFold 3 via GPU-accelerated MSA construction

Perry, B. C.; Kim, J.; Romero, P. A.

2026-02-18 biochemistry
10.64898/2026.02.17.706409 bioRxiv
Show abstract

AlphaFold 3 (AF3) enables accurate biomolecular modeling but is limited by slow, CPU-bound multiple sequence alignment (MSA) generation. We introduce AlphaFast, a drop-in framework that integrates GPU-accelerated MMseqs2 sequence search to remove this bottleneck. AlphaFast achieves a 68.5x speedup in MSA construction and a 22.8x reduction in end-to-end runtime on a single GPU, and delivers predictions in 8 seconds per input on four GPUs while maintaining indistinguishable structural accuracy. A serverless deployment enables structure prediction for as little as $0.035 per input. Code is available at https://github.com/RomeroLab/alphafast

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

1
Nature Methods
○ 385 papers in training set
Top 0.2%
26.1%
Similar papers in this journal
2
Nature Communications
○ 5641 papers in training set
Top 16%
11.7%
Similar papers in this journal
3
Bioinformatics
◐ 1204 papers in training set
Top 3%
9.5%
Similar papers in this journal
4
Protein Science
○ 246 papers in training set
Top 0.6%
6.6%
Similar papers in this journal
50% of probability mass above
5
Proceedings of the National Academy of Sciences
● 2444 papers in training set
Top 11%
4.4%
Similar papers in this journal
6
Nature Biotechnology
○ 172 papers in training set
Top 1%
3.2%
7
Journal of Molecular Biology
○ 232 papers in training set
Top 1%
2.6%
8
Journal of Chemical Information and Modeling
● 238 papers in training set
Top 1%
2.3%
9
Cell Reports Methods
○ 165 papers in training set
Top 1%
2.3%
10
Nucleic Acids Research
◐ 1281 papers in training set
Top 7%
2.3%
11
Communications Biology
○ 993 papers in training set
Top 9%
2.3%
12
Cell Systems
○ 201 papers in training set
Top 2%
2.1%
13
Genome Biology
○ 637 papers in training set
Top 7%
1.3%
14
Structure
○ 193 papers in training set
Top 2%
1.1%
15
Science
● 477 papers in training set
Top 7%
1.1%
16
Nature Computational Science
○ 55 papers in training set
Top 1%
1.1%
17
Advanced Science
○ 286 papers in training set
Top 7%
1.1%
18
PLOS Computational Biology
● 1863 papers in training set
Top 18%
1.0%
19
GigaScience
◐ 212 papers in training set
Top 4%
1.0%
20
SoftwareX
○ 15 papers in training set
Top 0.2%
1.0%
21
eLife
● 5828 papers in training set
Top 67%
0.8%
22
Communications Chemistry
○ 48 papers in training set
Top 1%
0.8%
23
iScience
○ 1154 papers in training set
Top 37%
0.8%
24
Scientific Reports
○ 3612 papers in training set
Top 78%
0.6%
25
PLOS ONE
● 5266 papers in training set
Top 63%
0.6%
26
Angewandte Chemie International Edition
○ 93 papers in training set
Top 2%
0.6%

"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.