“EnzyREx: deep learning enables reaction-based comprehensive enzyme screening” (2026), by Keisuke Hirota et al. Genetics and Biochemistry research paper. Original abstract below, unchanged. A free PDF is available.

EnzyREx: deep learning enables reaction-based comprehensive enzyme screening

Keisuke Hirota, Takuji Yamada · November 3, 2026 · Genetics and Biochemistry

Abstract

Orphan enzymes for which the corresponding genes remain unidentified represent a major gap in functional annotation in genomic analysis. We present EnzyREx, an AI-based framework for comprehensive enzyme screening, to explore novel enzyme candidates. EnzyREx consists of two deep learning models to classify proteins as enzymes and non-enzymes and predict catalytic capability for enzyme‒reaction pairs. Both models achieved performance comparable to or better than that of existing software. When applied to 1,255 microbiome-related orphan reactions, EnzyREx identified candidate proteins for 897 reactions. EnzyREx provides a scalable approach for linking unannotated proteins to orphan reactions and extending functional annotation.

Free PDF DOI: 10.1186/s13059-026-04312-7