Language models can identify enzymatic binding sites in protein sequences

Yves Gaetan Nana Teukam; Loïc Kwate Dassi; Matteo Manica; Daniel Probst; Philippe Schwaller; Teodoro Laino

doi:10.1016/j.csbj.2024.04.012

Language models can identify enzymatic binding sites in protein sequences

Comput Struct Biotechnol J. 2024 Apr 30:23:1929-1937. doi: 10.1016/j.csbj.2024.04.012. eCollection 2024 Dec.

Authors

Yves Gaetan Nana Teukam¹, Loïc Kwate Dassi¹, Matteo Manica¹, Daniel Probst^{1

2}, Philippe Schwaller^{1

2}, Teodoro Laino^{1

2}

Affiliations

¹ IBM Research Europe, Saümerstrasse 4, 8803 Rüschlikon, Switzerland.
² National Center for Competence in Research-Catalysis (NCCR-Catalysis), Switzerland.

Abstract

Recent advances in language modeling have had a tremendous impact on how we handle sequential data in science. Language architectures have emerged as a hotbed of innovation and creativity in natural language processing over the last decade, and have since gained prominence in modeling proteins and chemical processes, elucidating structural relationships from textual/sequential data. Surprisingly, some of these relationships refer to three-dimensional structural features, raising important questions on the dimensionality of the information encoded within sequential data. Here, we demonstrate that the unsupervised use of a language model architecture to a language representation of bio-catalyzed chemical reactions can capture the signal at the base of the substrate-binding site atomic interactions. This allows us to identify the three-dimensional binding site position in unknown protein sequences. The language representation comprises a reaction-simplified molecular-input line-entry system (SMILES) for substrate and products, and amino acid sequence information for the enzyme. This approach can recover, with no supervision, 52.13% of the binding site when considering co-crystallized substrate-enzyme structures as ground truth, vastly outperforming other attention-based models.