/tools

PepToCodes is a Python script designed to convert amino acid SMILES representations into one-letter or three-letter codes for further analysis. It supports both a native database of 20 amino acids and a larger Norine database for more comprehensive peptide analysis.

The Peptaloid Database is a comprehensive platform for exploring peptide alkaloid molecules, offering advanced search capabilities, molecular data visualization, and access to extensive datasets. It facilitates drug discovery by providing detailed information on over 161,000 peptide alkaloid compounds, including their structural and physicochemical properties.

PepDesign is a modular pipeline designed for computational peptide design and optimization. It integrates advanced deep learning models for generating peptide sequences, predicting structures, and scoring based on physicochemical properties, making it a valuable tool in drug discovery and protein engineering.

The 'library-analysis' tool is designed for analyzing peptide libraries and performing quantitative structure-activity relationship (QSAR) modeling. It processes FASTQ files to generate various outputs related to peptide composition and scoring.

The 'pu-peptides' repository develops deep learning models to predict various properties of peptides using only positive examples. It addresses the challenge of limited negative data in peptide databases by employing positive-unlabeled learning to infer properties such as solubility and binding affinity.

gmx_wheel+ is a tool designed for analyzing and visualizing the orientation of helical antimicrobial peptides derived from molecular dynamics simulations. It facilitates the design process by providing advanced projections and analysis of peptide structures, aiding in the development of new analogs.

This project involves the computational design of a multi-epitope vaccine against the Monkeypox virus, utilizing immunoinformatics and molecular docking techniques to predict immunogenic epitopes and evaluate their interactions with immune receptors.

This repository provides source code and tutorials for designing D-peptides that target the SARS-CoV-2 Receptor Binding Domain and Main Protease. It includes tools for scaffold generation and docking processes, making it a useful resource for molecular design and simulation in the context of drug discovery.

The AMP-Modification pipeline is designed to modify antimicrobial peptides (AMPs) using a genetic algorithm. It employs a fine-tuned protein language model to filter and select active peptides based on their physicochemical properties and structural constraints.

Molcraft provides minimalistic and powerful implementations of graph neural networks tailored for molecular machine learning. It includes features for predicting molecular properties and supports various molecule types, including peptides, through context-aware models.

Munis is a tool designed for predicting peptide-MHC interactions using machine learning models. It allows users to run predictions on protein sequences and train new models based on provided datasets.

HPL-APMS is a framework for predicting peptide binding to HLA alleles and automating the design of high-affinity peptides. It utilizes hierarchical progressive learning and protein language models to enhance predictions for unseen alleles, making it a valuable tool in immunotherapy research.

The solvMPCONF196 repository provides geometries of 196 uncharged conformers of 13 peptides and macrocycles, solvated with explicit water molecules. This dataset serves as a benchmark for evaluating molecular simulations and property predictions in computational chemistry.

PEPSGO is a tool for predicting the three-dimensional structure of peptides from their amino acid sequences using a custom global optimization algorithm. It utilizes Rosetta fragments and a multivariate quantile function to explore the search space for accurate structure prediction.

synth-pdb is a command-line tool designed to generate realistic Protein Data Bank (PDB) files with customizable sequences and conformations for proteins and peptides. It serves as a resource for testing, benchmarking, and educational purposes in structural biology and bioinformatics.

This tool is designed for de novo peptide drug design targeting beta-lactamase using deep learning techniques. It validates the designed peptides through structure prediction, docking, and molecular dynamics simulations, contributing to antimicrobial resistance (AMR) drug discovery.

labmate-mcp is an AI lab companion that offers 81 tools for literature search, compound synthesis, analysis, and publication writing. It includes functionalities for predicting molecular properties, planning syntheses, and generating molecular data, making it a valuable resource for researchers in computational chemistry and molecular biology.

The MHC1-Binding repository contains a Jupyter Notebook that builds classifiers to predict if specific peptides will be presented by MHC1 proteins based on their allele names. It utilizes various machine learning and deep learning methods to analyze and process peptide data, contributing to the understanding of immune responses and potential immunotherapy development.