Aligning sequence and structure representations leveraging protein domains for function prediction

doi:10.1016/j.eswa.2025.127246

Aligning sequence and structure representations leveraging protein domains for function prediction

2025-06-01

0

PRE

AI

Athanasios V. Vasilakos

Z

Zhixiang Ren *

DOI：10.1016/j.eswa.2025.127246

Original

Original request for help

Save

Abstract

En 中文

Protein function prediction is traditionally approached through sequence or structural modeling, often neglecting the effective fusion of diverse data sources. Protein domains, as functionally independent building blocks, determine a protein's biological function, yet their potential has not been fully exploited in function prediction tasks. To address this, we introduce a modality-fused neural network leveraging function-aware domain embeddings as a bridge. We pre-train these embeddings by aligning domain semantics with Gene Ontology (GO) terms and textual descriptions. Additionally, we partition proteins into sub-views based on continuous domain regions for contrastive learning, supervised by a novel triplet InfoNCE loss. Our method outperforms state-of-the-art approaches across various benchmarks, and clearly differentiates proteins carrying distinct functions compared to the competitor.

Keywords:

Protein function prediction

Protein domain

Deep learning

Functional priors

Contrastive learning

Aligning sequence and structure representations leveraging protein domains for function prediction

Abstract

Journal

Organization