# DPI-score: a deep learning-based metric for assessing protein-protein interfaces in cryoEM derived assemblies

Source: [nature.com](https://www.nature.com/articles/s41467-026-79090-8) · 2026-11-14 · Life
Page: https://themorning.science/story/dpi-score-a-deep-learning-based-metric-for-assessing-protein-protein-interfaces-a8f121783eb692cb
Kind: research · Impact: medium

## In brief

- DPI-score is a deep learning-based metric for assessing protein-protein interfaces in cryoEM-derived assemblies.
- The score uses only structural coordinates, without engineered interface features.
- The study was published in Nature Communications on 14 November 2026 (doi:10.1038/s41467-026-79090-8).

## Why it matters

Structural biologists building models of multi-protein complexes from cryoEM maps need reliable ways to judge whether the contacts between chains are credible; a coordinate-only deep learning metric offers a way to check these interfaces.

## Timeline

- 2026-11-14: Study published online in Nature Communications

## Coverage

- [DPI-score: a deep learning-based metric for assessing protein-protein interfaces in cryoEM derived assemblies](https://www.nature.com/articles/s41467-026-79090-8) — nature.com, 2026-11-14

Tags: cryoem, protein-protein interfaces, deep learning, structural biology, model validation

---

Brief generated by AI (claude-haiku-5-5) from the reporting listed under Coverage. Headlines belong to their publishers. The Morning Science – https://themorning.science