Conversion between IUPAC names and SMILES chemical structures using context free grammars.

No-frills command-line SMARTS substructure search for DSSTOX

⬢⬢⬢ Organizing and processing tables of chemical structures.

One-hot encoding for simple molecular-input line-entry system (SMILES) strings

Generate synthetic chemical reactions from a retrosynthetic SMARTS template and a stock of molecules.

GNN powered visualize solutions to SMILE(simplified Molecularinputline entrysystem)

ChemBERTa + RL pipeline for de novo EGFR inhibitor design with docking validation.

Python tool for scraping geometric X-ray Data from the Cambridge Structural Database

This project trains a Morgan Fingerprint model to predict lipophilicity.

A Python library for efficient manipulation, conformer optimization, and molecular structure data.

An MCP bundle toolkit for manipulating and interpreting SMILES strings

Многомодальная нейросетевая система для предсказания биоактивности молекул и генерации новых структур с заданными свойствами. Проект разработан для поиска потенциальных сокристаллов теофиллина с использованием генеративных моделей и эволюционной оптимизации.

A Python Framework for Integrating Classical and Quantum Mechanical Simulation Data, Experimental Data, and Cheminformatics into a Unified Predictive Engine for Macromolecular Design

Research-oriented Streamlit GUI for single-step retrosynthesis from SMILES: PyTorch-based inference, RDKit validation & depiction, dataset-driven model checkpoints and extensible decoding.

A modular drug-design AI Agent

Semi-supervised learning techniques (pseudo-label, mixmatch, and co-training) for pre-trained BERT language model amidst low-data regime based on molecular SMILES from the Molecule Net benchmark.

Here I provide a script that allows the user to browse molport pages and collect SMILES of molecules of their interest

ModChemBERT: ModernBERT as a Chemical Language Model

Program to survey all possible molecular structures of fragment ion based on exact-mass and molecular formula of it. Ideal for ESI-Tandem MS/MS fragment's structure studies.

BioInsectiNet is a Python-based framework that uses neural networks and genetic algorithms to design and optimize bioinsecticides targeting specific proteins. It generates novel SMILES compounds, predicts toxicity and affinity using IC50 data from FASTA sequences, and performs molecular docking for 3D interaction analysis.