Grouped Neutralization Learning Predicts Antibody Sensitivity
Researchers publish a March 23, 2026 PLOS Computational Biology paper presenting grouped neutralization learning (GNL), a method that predicts virus-antibody neutralization values from viral genetic sequences and antibody similarity. The approach integrates sequence-based prediction, antibody-level information sharing, and low-rank matrix refinement, showing improved accuracy and interpretability across large HIV-1 neutralization datasets, and estimating sensitivity for unmeasured viral sequences.
Scoring Rationale
New interpretable method with peer-reviewed validation and usable code; scope remains focused on HIV and similar pathogens.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
