Accurate Prediction of Ligand-Protein Interaction Affinities with Fine-Tuned Small Language Models

TL;DR

This paper presents a novel approach using instruction fine-tuned small language models to accurately predict ligand-protein interaction affinities, outperforming traditional ML and FEP+ methods in zero-shot settings for drug discovery.

Contribution

It introduces a new method leveraging fine-tuned generative small language models for ligand-protein affinity prediction using only SMILES and amino acid sequences.

Findings

Outperforms ML and FEP+ methods in accuracy

Effective in zero-shot prediction scenarios

Applicable to challenging therapeutic targets

Abstract

We describe the accurate prediction of ligand-protein interaction (LPI) affinities, also known as drug-target interactions (DTI), with instruction fine-tuned pretrained generative small language models (SLMs). We achieved accurate predictions for a range of affinity values associated with ligand-protein interactions on out-of-sample data in a zero-shot setting. Only the SMILES string of the ligand and the amino acid sequence of the protein were used as the model inputs. Our results demonstrate a clear improvement over machine learning (ML) and free-energy perturbation (FEP+) based methods in accurately predicting a range of ligand-protein interaction affinities, which can be leveraged to further accelerate drug discovery campaigns against challenging therapeutic targets.