Skip to content
DraconiticResearch
Search Sources API
197,785 studies31 sourcesUpdated 15 min ago
Peer-reviewedCC BYFundamental Scientific Reports in Multidisciplinary Areas

GastroDL-Fusion: A Dual-Modal Deep Learning Framework Integrating Protein–Ligand Complexes and Gene Sequences for Gastrointestinal Disease Drug Discovery

  • Ziyang Gao1
  • Annie Cheung2
  • Yihao Ou3
  1. 1Technical University of Munich, Munich, German
  2. 2University of Michigan, Ann Arbor, MI, USA
  3. 3Georgia Institute of Technology, Atlanta, GA, USA
Open at sourceDOICite

Abstract

Accurate prediction of protein–ligand binding affinity plays a pivotal role in accelerating the discovery of novel drugs and vaccines, particularly for gastrointestinal (GI) diseases such as gastric ulcers, Crohn’s disease, and ulcerative colitis. Traditional computational models often rely on structural information alone and thus fail to capture the genetic determinants that influence disease mechanisms and therapeutic responses. To address this gap, we propose GastroDL-Fusion, a dual-modal deep learning framework that integrates protein–ligand complex data with disease-associated gene sequence information for drug and vaccine development. In our approach, protein–ligand complexes are represented as molecular graphs and modeled using a Graph Isomorphism Network (GIN), while gene sequences are encoded into biologically meaningful embeddings via a pre-trained Transformer (ProtBERT/ESM). These complementary modalities are fused through a multi-layer perceptron to enable robust cross-modal interaction learning.We evaluate the model on benchmark datasets of GI disease-related targets, demonstrating that GastroDL-Fusion significantly improves predictive performance over conventional methods. Specifically, the model achieves a mean absolute error (MAE) of 1.12 and a root mean square error (RMSE) of 1.75, outperforming CNN, BiLSTM, GIN, and Transformer-only baselines. These results confirm that incorporating both structural and genetic features yields more accurate predictions of binding affinities, providing a reliable computational tool for accelerating the design of targeted therapies and vaccines in the context of gastrointestinal diseases.

Related

  1. Building the Data Foundations for the Design and Discovery of Transition Metal Complexes
    Hannes Kneiding, Bastian Bjerkem Skjelstad, Luca Cibinel +3
    ChemRxivPreprintCC BYFull textMaterials & polymersChemRxiv
  2. Integrating artificial intelligence and molecular simulation for peptide discovery: limitations, validation and decision-centered workflows
    Xu Wu, Wei Fu
    ChemRxivPreprintLicence not statedGeneral chemistryChemRxiv
  3. Integrating Artificial Intelligence Into Drug Discovery From Medicinal Plants: Current Applications and Infrastructural Challenges
    Amit Gangwal, Azim Ansari, Mohd Usman Mohd Siddique +3
    Chemical Biology & Drug DesignPeer-reviewedCC BY6 citedMedicinal chemistry & drug discoveryOpenAlex
  4. TKitOC PRO v8+: An Interactive Cloud-Based Platform Integrating Target Discovery and Chemoinformatics Workflows
    Angel Eduardo Nava-Saldivar, Jonnathan Guadalupe Santillán-Benítez
    ChemRxivPreprintCC BY-NC-NDFull textGeneral chemistryChemRxiv
  5. Decision Framework for Polymer Composite Linings in Trenchless Renewal: Integrating Materials Durability, Operational Performance, Application Quality, and Environmental Sustainability
    Parastou Kharazmi
    ChemRxivPreprintLicence not statedGeneral chemistryChemRxiv

Cite

BibTeXRIS
BibTeX
@article{gao2026gastrodlfusion,
  title = {{GastroDL-Fusion: A Dual-Modal Deep Learning Framework Integrating Protein–Ligand Complexes and Gene Sequences for Gastrointestinal Disease Drug Discovery}},
  author = {Ziyang Gao and Annie Cheung and Yihao Ou},
  journal = {Fundamental Scientific Reports in Multidisciplinary Areas},
  year = {2026},
  month = may,
  doi = {10.66238/fsrma80},
  url = {https://doi.org/10.66238/fsrma80},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - GastroDL-Fusion: A Dual-Modal Deep Learning Framework Integrating Protein–Ligand Complexes and Gene Sequences for Gastrointestinal Disease Drug Discovery
AU  - Ziyang Gao
AU  - Annie Cheung
AU  - Yihao Ou
PY  - 2026
DA  - 2026/05/25
JO  - Fundamental Scientific Reports in Multidisciplinary Areas
DO  - 10.66238/fsrma80
UR  - https://doi.org/10.66238/fsrma80
AB  - Accurate prediction of protein–ligand binding affinity plays a pivotal role in accelerating the discovery of novel drugs and vaccines, particularly for gastrointestinal (GI) diseases such as gastric ulcers, Crohn’s disease, and ulcerative colitis. Traditional computational models often rely on structural information alone and thus fail to capture the genetic determinants that influence disease mechanisms and therapeutic responses. To address this gap, we propose GastroDL-Fusion, a dual-modal deep learning framework that integrates protein–ligand complex data with disease-associated gene sequence information for drug and vaccine development. In our approach, protein–ligand complexes are represented as molecular graphs and modeled using a Graph Isomorphism Network (GIN), while gene sequences are encoded into biologically meaningful embeddings via a pre-trained Transformer (ProtBERT/ESM). These complementary modalities are fused through a multi-layer perceptron to enable robust cross-modal interaction learning.We evaluate the model on benchmark datasets of GI disease-related targets, demonstrating that GastroDL-Fusion significantly improves predictive performance over conventional methods. Specifically, the model achieves a mean absolute error (MAE) of 1.12 and a root mean square error (RMSE) of 1.75, outperforming CNN, BiLSTM, GIN, and Transformer-only baselines. These results confirm that incorporating both structural and genetic features yields more accurate predictions of binding affinities, providing a reliable computational tool for accelerating the design of targeted therapies and vaccines in the context of gastrointestinal diseases.
SN  - 3134-7509
SN  - 3134-7517
ER  -