/tools

The `proteinogenic` library generates chemical structures for protein sequences in the form of SMILES strings. It allows users to encode amino acid sequences and create in-memory representations of protein formulas, making it a useful tool for bioinformatics and cheminformatics applications.

ChemDraw Server is a FastAPI-based API service that facilitates the conversion between chemical names and SMILES strings, as well as the conversion of SMILES to RDKit molecule objects. It is designed for integration into chemical drawing tools and automation pipelines, making it a useful tool in the field of cheminformatics.

This repository offers a Docker deployment for a Flask application that facilitates the generation of protein embeddings using pretrained models from the TAPE project. It also includes capabilities for visualizing these embeddings through PCA, making it a useful tool for protein analysis and representation.

The molecule-ELECTRA repository provides a framework for pre-training and evaluating advanced neural network architectures on molecular data. It utilizes the ELECTRA method to enhance the understanding and prediction of molecular properties through machine learning techniques.

The OPSIN-ws Docker image provides a web service for converting systematic IUPAC chemical names into various structure formats. It facilitates the interpretation of chemical nomenclature, making it useful for cheminformatics applications.

PyMolinfo is a Python package designed for advanced molecular analysis by converting molecular structures into graph representations. It allows researchers to load various molecular file formats, analyze functional groups, and compute distances and angles within molecular graphs.

Drug-BERT is a tool designed for pre-training representations of drug sub-structures to enhance the prediction of molecular properties. It aims to improve the accuracy of property predictions in drug discovery by leveraging advanced machine learning techniques.

Enumeration-aware Molecular Transformers provides a suite of neural language model tools designed for pre-training and fine-tuning SMILES-based molecular language models. It incorporates techniques like contrastive learning and multitask regression to enhance molecular representation learning.

The Wiswesser Line Notation Project offers a chemical converter and extractor for Wiswesser Line Notation (WLN), allowing users to read and write WLN to various chemical formats. It also includes a tool for extracting WLN sequences from documents, enhancing the accessibility of chemical information.

cov2vec is a tool designed to obtain embeddings of the SARS CoV-2 genome by encoding viral genomes with protein language models. It facilitates the analysis of mutations and their effects on protein sequences, contributing to biomedical machine learning tasks.

Pymatgen (Python Materials Genomics) is an open-source Python library designed for materials analysis, providing flexible classes for representing elements, sites, molecules, and structures. It supports various electronic structure codes and offers powerful analysis tools for materials science research.

MMSchema is a vocabulary specification for molecular mechanics that supports various file formats like JSON, YAML, and HDF5. It facilitates the representation of molecular data, which can be useful in computational chemistry and related fields.

SuperSweetNet is a GNN model designed to learn latent space representations of glycans at an atomic level. This tool can be utilized in various applications related to molecular properties and design, particularly in the field of glycobiology.

The mmic_openff repository offers a translator for converting molecular representations between MMSchema and the OpenFF toolkit. It facilitates interoperability in molecular mechanics, allowing users to work with various molecular formats effectively.

Biosaic is a tokenizer library designed for DNA and protein sequences, enabling the conversion of sequences into k-mers and embeddings. It supports model training use-cases with state-of-the-art encoding architectures inspired by AlphaFold-2.

The mmic_qcschema repository provides a converter for translating molecules between MMSchema and QCSchema formats. This tool facilitates interoperability in computational molecular science, aiding in the representation and manipulation of molecular data.