Paraplume Predicts Antibody Paratopes At Scale
Researchers Athènes, Woolfe, Mora, and Walczak publish Feb. 18, 2026 in PLoS Computational Biology a sequence-based paratope predictor, Paraplume, that uses protein language model embeddings to identify antibody binding residues without structural input. Paraplume outperforms existing methods on multiple benchmarks, is available as a PyPI package and GitHub repository, and shows antigen-driven somatic hypermutation correlates with increased paratope size across repertoires.
Key Points
- 1Demonstrates Paraplume predicts paratopes from antibody sequences using PLM embeddings, outperforming structural methods
- 2Avoids 3D structure requirements, enabling scalable repertoire-wide paratope mapping and faster therapeutic screening
- 3Provides PyPI/GitHub implementation and embeddings reweighting improving binder classification and epitope binning
Scoring Rationale
High methodological novelty and peer-reviewed validation; impact focused on antibody paratope prediction rather than broad AI infrastructure.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
