Publications

---

Foutch, D., Pham, B., & Shen, T. Y. (2021). Protein conformational switch discerned via network centrality properties. Computational and Structural Biotechnology Journal, 19, 3599-3608. https://doi.org/10.1016/j.csbj.2021.06.004 PubMed+1

Pham, B., Cheng, Z., Lopez, D., Lindsay, R. J., Foutch, D., Majors, R. T., & Shen, T. Y. (2022). Statistical analysis of protein–ligand interaction patterns in nuclear receptor RORγ. Frontiers in Molecular Biosciences, 9, Article 904445. https://doi.org/10.3389/fmolb.2022.904445 Frontiers+1

Haratipour, Z., Foutch, D., & Blind, R. D. (2024). A novel heuristic of rigid docking scores positively correlates with full-length nuclear receptor LRH-1 regulation. Computational and Structural Biotechnology Journal, 23, 3065-3080. https://doi.org/10.1016/j.csbj.2024.07.021 PubMed+1

Foutch, D. & Blind, R. D. (2025). PDB2Graph: A PyMOL Plugin for Protein Structure Network Construction and Visualization (Version 1.0) [Software]. https://github.com/…

Title:  PDB2Graph: A PyMOL Plugin for Protein Structure Network Analysis

Venue:  Blind Lab, Department of Biochemistry, Vanderbilt University

Summary:  Delivered a live demonstration of PDB2Graph, a PyMOL plugin that allows researchers to generate protein structure networks (PSNs) directly from structural data, compute four centrality metrics, and analyze residue-level interaction patterns. The session included real-time comparison of PSNs across identical LRH-1 structures bound to different ligands, illustrating how centrality shifts reveal conformational changes and allosteric communication pathways.

Materials:
• Slides (PDF)
• Repository (GitHub)

Title: Protein Structure Networks and Graph Neural Networks for Allosteric State Classification

Venue: Stanford University (Professional Certificate Program)

Summary: A live presentation demonstrating a Graph Attention Network (GAT) trained on protein structure networks to distinguish high- and low-energy conformations of the nuclear receptor LRH-1. The talk included visualizations of multi-head attention mechanisms and highlighted the workflow being executed entirely on a custom-built dual-Xeon AI workstation designed for graph neural network research.

Materials:
• Slides (PDF)
• Repository (GitHub)

XCS224W–Machine Learning with Graphs (Completed, June 2025)
Credential ID: X844665
View Certificate

Graph Neural Networks (GCN, GAT, GraphSAGE), Graph ML pipelines, link prediction, graph classification, message-passing theory, and large-scale graph workflows.

XCS224N–Natural Language Processing with Deep Learning (In Progress, 2025)
 

Dependency parsing, RNNs/LSTMs, word embeddings, sequence-to-sequence models, attention, transformer architectures, neural machine translation.

XCS236–Deep Generative Models (Jan 26-Apr 5, 2026)

Variational autoencoders, diffusion models, normalizing flows, energy-based models, latent-variable modeling, probabilistic generative modeling for high-dimensional data.