Repurposing PeTriBERT for Protein Protein Interaction Prediction with Sequence Structure Fusion
Abstract
Recent advances in artificial intelligence have enabled models to capture protein sequence and structural features. We present PeTriPPI2, a hybrid framework for protein--protein interaction (PPI) prediction that combines sequence embeddings from ESM-2 with structural representations from PeTriBERT, an encoder originally designed for inverse folding. We evaluate this sequence--structure approach on the Pinder dataset. On a test set of 2,342 protein pairs, the selected checkpoint achieves 0.898 accuracy, 0.917 precision, 0.876 recall, and 0.896 F1 score for the interacting class (AUROC 0.963). Compared with SpatialPPIv2 on the same test data, PeTriPPI2 has higher precision (0.917 versus 0.886) but lower recall (0.876 versus 0.942) and F1 score (0.896 versus 0.913). Component ablations indicate that both the sequence and structural pathways contribute to this configuration.
Related
- From single-sequence structure prediction to protein fitness landscape through a composable, epistasis-aware mutation atlas
- Sequence-only prediction of VHH nanobody epitopes and paratopes on HER2 with a frozen protein language model ( ESM2) and a structure-derived residue prior
- A high-resolution protein-RNA binding atlas for structure-aware interaction prediction and post-transcriptional regulatory modeling
- GeoEPred: A multimodal structure-aware geometric deep learning framework for Gram-negative bacterial secreted effector prediction with sequence semantics
- Programmable Biochemical Logic Computation Through Repurposing Natural bZIP Protein Interaction Networks
BibTeX
@misc{wavreille2026repurposing,
title = {{Repurposing PeTriBERT for Protein Protein Interaction Prediction with Sequence Structure Fusion}},
author = {Arthur Wavreille and Gabriel Krouk and Baldwin Dumortier},
howpublished = {HAL},
year = {2026},
month = sep,
doi = {10.64898/2026.09.29.755338},
url = {https://hal.science/hal-05774253v1},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Repurposing PeTriBERT for Protein Protein Interaction Prediction with Sequence Structure Fusion
AU - Arthur Wavreille
AU - Gabriel Krouk
AU - Baldwin Dumortier
PY - 2026
DA - 2026/09/29
PB - HAL
DO - 10.64898/2026.09.29.755338
UR - https://hal.science/hal-05774253v1
AB - Recent advances in artificial intelligence have enabled models to capture protein sequence and structural features. We present PeTriPPI2, a hybrid framework for protein--protein interaction (PPI) prediction that combines sequence embeddings from ESM-2 with structural representations from PeTriBERT, an encoder originally designed for inverse folding. We evaluate this sequence--structure approach on the Pinder dataset. On a test set of 2,342 protein pairs, the selected checkpoint achieves 0.898 accuracy, 0.917 precision, 0.876 recall, and 0.896 F1 score for the interacting class (AUROC 0.963). Compared with SpatialPPIv2 on the same test data, PeTriPPI2 has higher precision (0.917 versus 0.886) but lower recall (0.876 versus 0.942) and F1 score (0.896 versus 0.913). Component ablations indicate that both the sequence and structural pathways contribute to this configuration.
ER -