title | layout | nav_order |
---|---|---|
Overview |
home |
1 |
Opensource Toolkits for Molecular Science
OpenAD is an open-source framework developed by IBM Research, aggregating a number of molecular science toolkits into a single API that can be accessed by command line, a Jupyter Notebook and (soon) an API.
The goal of openAD is to provide a common language for scientists to interact with a multitude of of molecular tools to simplify the triage process and drastically accelerate your development timelines.
The OpenAD Beta integrates with the following IBM Research [toolkits]({% link base-concepts.md %}#toolkits):
-
Deep Search for Scientific Discovery
Connecting & ingesting unstructured data
DS4SD Docs {: .no-gap }More
The Deep Search toolkit uses AI to convert unstructured PDF documents into structured JSON files and enables you to automate knowledge extraction.
You can use it for both public and proprietary documents.
-
Computational Chemistry
RXN Video tutorials (account required) {: .no-gap }More
The Reaction toolkit uses AI to predict chemical reactions, retrosynthesis pathways and experimental procedures.
You can train AI models to build intelligence in your specific chemistry domain, and scale your analysis and model training while securing your data using features of the Discovery Platform.
-
Coming soon
Generative Toolkit for Scientific Discovery
Molecular modeling & inferences {: .no-gap }More
The Generative Toolkit accelerates hypothesis generation in the scientific discovery process. It provides a library for making state-of-the-art generative AI models easier to use. -
Coming soon
Simulation Toolkit for Scientific Discovery
Virtual Experiments {: .no-gap }More
The Simulation Toolkit simplifies the development, execution and dissemination of virtual experiments.
A virtual experiment is an application workflow which measures one or more characteristics of one or more input systems. It is the computational analog of a lab experiment.