Comprehensive Evaluation of Protein Language Model Embeddings for Drug-Target Affinity Prediction
- 1University of Belgrade - School of Electrical Engineering
Abstract
Accurate identification of drug-target interactions is consequential for novel drug discovery and development. Deep learning methods for drug-target affinity (DTA) prediction have shown great promise in accelerating drug discovery and reducing development costs. Although graph neural networks have improved drug representation learning for DTA prediction tasks, many models still struggle to effectively and efficiently capture protein information, limiting overall prediction accuracy. In this work, we systematically evaluate the impact of pre-trained protein language models (PLMs) on the downstream task of predicting binding affinity between drugs and target proteins. We design multiple experiments across four different molecular representation backbones and assess the effect of incorporating PLM embeddings, comparing their performance to classical 1D convolution methods. We evaluate four families of PLMs which we integrate into PLM-GraphDTA, each built on distinct architectures and optimized for different tasks, including structure prediction, function prediction, and sequence unmasking. Additionally, we evaluate DeepGraphDTA, an architectural modification of the baseline convolution method designed to improve protein representation learning. The models are evaluated on two benchmark datasets, Davis and KIBA, using concordance index (CI) and mean squared error (MSE) as performance metrics. We further evaluate the generalization power of each model using cold-start train and test splits, and analyze the per-protein contribution to total CI. The results indicate simple architectural modifications to traditional convolution methods may be sufficient to bridge the gap to large pre-trained PLMs.
Related
- Benchmarking Large Language Models for Assay-Aware Drug-Target Binding Affinity Prediction
- Antibody-Antigen Affinity Prediction with Chain-Aware Protein Language Modeling
- Comprehensive evaluation of antibody language models across multiple antibody design tasks
- Ground-truth evaluation of sparse autoencoder features on spatial, single-cell and language-model embeddings
- A Context-Aware 14-3-3 Binding Predictor Enabled by Data Augmentation, Multi-Scale Biological Features, and Protein Language Model Embeddings
BibTeX
@misc{marijan2026comprehensive,
title = {{Comprehensive Evaluation of Protein Language Model Embeddings for Drug-Target Affinity Prediction}},
author = {M. Marijan and I. Tanasijevic},
howpublished = {bioRxiv},
year = {2026},
month = aug,
doi = {10.64898/2026.08.31.748056},
url = {https://www.biorxiv.org/content/10.64898/2026.08.31.748056v1},
note = {CC BY-NC},
}
RIS
TY - UNPB
TI - Comprehensive Evaluation of Protein Language Model Embeddings for Drug-Target Affinity Prediction
AU - M. Marijan
AU - I. Tanasijevic
PY - 2026
DA - 2026/08/31
PB - bioRxiv
DO - 10.64898/2026.08.31.748056
UR - https://www.biorxiv.org/content/10.64898/2026.08.31.748056v1
AB - Accurate identification of drug-target interactions is consequential for novel drug discovery and development. Deep learning methods for drug-target affinity (DTA) prediction have shown great promise in accelerating drug discovery and reducing development costs. Although graph neural networks have improved drug representation learning for DTA prediction tasks, many models still struggle to effectively and efficiently capture protein information, limiting overall prediction accuracy. In this work, we systematically evaluate the impact of pre-trained protein language models (PLMs) on the downstream task of predicting binding affinity between drugs and target proteins. We design multiple experiments across four different molecular representation backbones and assess the effect of incorporating PLM embeddings, comparing their performance to classical 1D convolution methods. We evaluate four families of PLMs which we integrate into PLM-GraphDTA, each built on distinct architectures and optimized for different tasks, including structure prediction, function prediction, and sequence unmasking. Additionally, we evaluate DeepGraphDTA, an architectural modification of the baseline convolution method designed to improve protein representation learning. The models are evaluated on two benchmark datasets, Davis and KIBA, using concordance index (CI) and mean squared error (MSE) as performance metrics. We further evaluate the generalization power of each model using cold-start train and test splits, and analyze the per-protein contribution to total CI. The results indicate simple architectural modifications to traditional convolution methods may be sufficient to bridge the gap to large pre-trained PLMs.
ER -