AlphaInterp: Revealing how AlphaFold 3 processes structure, confidence, and evolutionary information

How does evolutionary information become structural knowledge? AlphaFold 3 predicts protein structures and biomolecular assemblies with remarkable accuracy, yet how it transforms comparative sequence information into three-dimensional structure remains poorly understood. Here, we use mechanistic interpretability to follow this transformation through AlphaFold 3 and identify a computational principle underlying its structural inference. By extracting single- and pair-representations across the model’s forward pass and combining representational-geometry analysis, linear probing, activation patching, and perturbation of the multiple-sequence alignment (MSA) and its phylogenetic diversity, we show that AlphaFold 3 does not simply infer structure from raw sequence nor from MSA depth. Instead, a small number of sufficiently divergent homologues provides enough comparative information to largely recover the structure predicted using the full MSA. The pairformer transforms this evolutionary signal into a compact latent space in which residue-pair distances, contacts, fold similarity, and structural confidence become progressively decodable. Removing the MSA disrupts these latent geometries. Replacing it with evolutionarily similar homologues provides substantially less information, while evolutionarily unrelated sequences fail to rescue prediction despite preserving the formal alignment input. In contrast, phylogenetically diverse MSA subsampling largely restores both prediction accuracy and representational coherence. These results suggest that AlphaFold 3 has learned to infer structure by constraining its geometric priors and structural hypotheses by the sequence variation of the MSA. This framework helps explain both the model’s remarkable predictive power and its characteristic failure modes, with direct implications for protein design, conformational diversity, out-of-distribution generalization, and the interpretation of model confidence.

Authors

Institutions

Publication Details

Journal
PLoS Computational Biology
Published
2026-11-19
DOI
https://doi.org/10.1371/journal.pcbi.1014981
Citations
11
Primary Topic
Protein Structure and Dynamics
Type
article
Field-Weighted Citation Impact
12.31

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

AlphaInterp: Revealing how AlphaFold 3 processes structure, confidence, and evolutionary information

Jeffrey Skolnick, Jonathan Feldman
11 citations
PLoS Computational Biology
Protein Structure and Dynamics
12.31
article

AlphaInterp: Revealing how AlphaFold 3 processes structure, confidence, and evolutionary information

Jeffrey Skolnick, Jonathan Feldman
article en
11 citations

Abstract

How does evolutionary information become structural knowledge? AlphaFold 3 predicts protein structures and biomolecular assemblies with remarkable accuracy, yet how it transforms comparative sequence information into three-dimensional structure remains poorly understood. Here, we use mechanistic interpretability to follow this transformation through AlphaFold 3 and identify a computational principle underlying its structural inference. By extracting single- and pair-representations across the model’s forward pass and combining representational-geometry analysis, linear probing, activation patching, and perturbation of the multiple-sequence alignment (MSA) and its phylogenetic diversity, we show that AlphaFold 3 does not simply infer structure from raw sequence nor from MSA depth. Instead, a small number of sufficiently divergent homologues provides enough comparative information to largely recover the structure predicted using the full MSA. The pairformer transforms this evolutionary signal into a compact latent space in which residue-pair distances, contacts, fold similarity, and structural confidence become progressively decodable. Removing the MSA disrupts these latent geometries. Replacing it with evolutionarily similar homologues provides substantially less information, while evolutionarily unrelated sequences fail to rescue prediction despite preserving the formal alignment input. In contrast, phylogenetically diverse MSA subsampling largely restores both prediction accuracy and representational coherence. These results suggest that AlphaFold 3 has learned to infer structure by constraining its geometric priors and structural hypotheses by the sequence variation of the MSA. This framework helps explain both the model’s remarkable predictive power and its characteristic failure modes, with direct implications for protein design, conformational diversity, out-of-distribution generalization, and the interpretation of model confidence.

PLoS Computational BiologyVol. 22(11)
Georgia Institute of Technology (US)
National Institute of General Medical Sciences
Openalex Percentile: Top 1%
Protein Structure and Dynamics
12.31
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon