Databases, Servers & Molecular Informatics Platforms

I have contributed as a developer, co-developer, and computational scientist to AI predictors, molecular databases, web servers, and cheminformatics platforms for peptide prediction, molecular interaction analysis, chemical-space exploration, and structure-based drug discovery. These platforms highlight a broader effort to build open, reproducible, and data-driven computational resources that connect molecular modeling, bioinformatics, non-covalent interactions, peptide science, and therapeutic discovery.

Digital research infrastructure

Computational resources for peptide AI, molecular informatics, and drug discovery

A curated collection of AI predictors, database resources, and molecular informatics tools developed for peptide prediction, chemical-space exploration, disease-focused discovery, and protein interaction analysis.

AI predictors MPDS resources Interaction databases

AI Predictor Developer

pLM4CPPs — Protein Language Model-Based Predictor for Cell-Penetrating Peptides

pLM4CPPs is a deep learning framework and web server for predicting cell-penetrating peptides. It uses pretrained protein language model embeddings with CNN-based classifiers to support CPP/non-CPP classification and prediction on new peptide sequences.

My contribution: Developed the prediction framework, model workflow, web-server concept, GitHub implementation, and reproducible notebooks for embedding generation, model training, evaluation, and new-sequence prediction.

Try server GitHub View article

Allergen Prediction Open Workflow

pLM4Alg — Protein Language Model Protocol for Allergen Prediction

pLM4Alg is a reproducible workflow for predicting allergenic proteins and peptides using ESM2 embeddings and CNN-based deep learning. It is designed for Google Colab execution and adaptation to related protein or peptide classification tasks.

My contribution: Developed the protocol structure, ESM2 embedding workflow, CNN pipeline, example datasets, prediction files, Google Colab notebooks, and GitHub documentation.

GitHub View chapter

Bioactive Peptides Open Workflow

UniDL4BioPep — Unified Deep Learning Workflow for Peptide Bioactivity Prediction

UniDL4BioPep is a unified deep learning workflow for predicting peptide bioactivity using protein language model-derived sequence representations. It supports transferable peptide modeling across multiple functional bioactivity classes.

My contribution: Documented and organized the protocol workflow for peptide bioactivity prediction using protein language models, deep learning classifiers, reproducible notebooks, and GitHub-accessible materials.

GitHub View chapter

Drug Discovery Portal MPDS Resource

MPDS COVID-19 — Disease-Specific Drug Discovery Portal

MPDS COVID-19 is an open-source disease-specific molecular informatics portal for COVID-19 drug discovery research. It integrates disease knowledge with molecular-property analysis, compounds, fragments, targets, pathways, and computational workflow modules.

My contribution: Contributed to the MPDS COVID-19 resource-development effort involving molecular informatics, module organization, computational drug discovery workflows, and disease-specific knowledge integration.

Visit portal View article

Drug Discovery Portal TB Informatics

MPDS TB — Molecular Property Diagnostic Suite for Tuberculosis

MPDS TB is a tuberculosis-focused molecular property and cheminformatics platform supporting compound analysis, property assessment, and structure-based therapeutic exploration for TB drug discovery research.

My contribution: Contributed to the MPDS-based tuberculosis molecular informatics resource for assessing molecular properties and therapeutic potential of compounds in TB-focused discovery workflows.

Visit portal View article

Fragment Library Drug Discovery

MPDS Fragment Library — Chemical-Space Exploration Resource

MPDS Fragment Library is a fragment-focused resource for fragment-based molecular design, chemical-space mapping, molecular property evaluation, and structure-oriented drug discovery workflows.

My contribution: Contributed to the development and publication of the MPDS fragment library module supporting systematic fragment analysis and chemical-space exploration.

Try server View article

Compound Library Chemical Space

MPDS Compound Library — Structure-Based Classification of Chemical Space

MPDS Compound Library is a compound-level chemical-space resource supporting compound organization, structural classification, molecular-property analysis, and structure-based exploration of molecular libraries.

My contribution: Contributed to compound-library organization and chemical-space classification within the MPDS molecular informatics framework.

Try server GitHub View article

Interaction Database Aromatic Interactions

A2ID 2.0 — Aromatic–Aromatic Interaction Database

A2ID 2.0 is a curated protein interaction database for systematic analysis of aromatic–aromatic interactions and π–π networks observed in experimentally determined protein structures.

My contribution: Contributed to the development and analysis of A2ID 2.0, supporting systematic investigation of aromatic residue networks and aromatic–aromatic interaction patterns in proteins.

Try server View article

Interaction Database Cation–Aromatic

CAD 2.0 — Cation–Aromatic Database

CAD 2.0 is a cation–aromatic interaction database for analyzing cation–π and related cation–aromatic motifs in experimentally determined protein structures.

My contribution: Contributed to database development and analysis for CAD 2.0, supporting systematic study of cation–aromatic interaction motifs across protein structures.

Try server View article

Development philosophy

My platform-development work focuses on building resources that are not only computationally useful, but also scientifically interpretable and reusable. Across peptide prediction, molecular interaction analysis, chemical-space exploration, and drug discovery workflows, I aim to connect curated data, molecular modeling, AI/ML methods, and open web-accessible tools for broader scientific use.