Atlas
Overview
Main Features
Quick Demo
Overview
The Atlas server is a high-performance search engine for large collections of chemical structures. It uses a thoroughly multi-threaded in-memory architecture and is meant to run on well equipped server hardware. It includes functionality to preprocess vendor catalogs of compounds in various formats. Its search functionality can be accessed via a REST API for easy integration into enterprise platforms. It also comes with a plugin for the DataWarrior application, which gives users easy access to the full Atlas server search functionality
Client-Server Architecture
Atlas consists of two components. A server component that runs compound searches and a client component that allows to draw structure queries, submit them to the server, receive results, and display, browse, and analyse them. The server component is implemented as Java REST based application that, when launched, reads all compound library files into memory and listens for incoming search requests.
Main Features
- Search by substructure including rich atom and bond query features, and exclude groups. - Search by chemical similarity, exact match, or non-stereo match. - Structure searches can be combined with vendor limitations and molecular weight ranges - Typical performance: millions searched and thousands of results retrieved within seconds - Input files can be sdf, sdf.gz, sdf.zip, dwar - must contain at least structures and company-IDs. - Native plugin for DataWarrior provides graphical dialog with structure editor to define searches; results appear in new DataWarrior window or are appended to previous search results.