ARTFEED — Contemporary Art Intelligence

Differentiable Graph Partitioning Interprets Protein Language Models

other · 2026-05-13

Researchers propose SoftBlobGIN, a framework that projects ESM-2 protein language model representations onto contact graphs for interpretable structural analysis. The method uses a Graph Isomorphism Network with differentiable Gumbel-softmax pooling to learn functional substructures. On enzyme classification tasks, it achieves 92.8% accuracy and 0.898 macro-F1. Unlike post hoc analysis, SoftBlobGIN produces directly auditable explanations, with GNNExplainer recovering biologically meaningful active-site residues and catalytic clusters.

Key facts

  • SoftBlobGIN is a plug-and-play framework for ESM-2 representations.
  • It projects representations onto protein contact graphs.
  • Uses a Graph Isomorphism Network with differentiable Gumbel-softmax pooling.
  • Achieves 92.8% accuracy and 0.898 macro-F1 on enzyme classification.
  • Produces directly auditable structural explanations.
  • GNNExplainer recovers active-site residues and functional clusters.
  • Framework is structure-aware and learns coarse functional substructures.
  • Addresses interpretability of dense latent spaces in protein language models.

Entities

Sources