An Intelligent Platform for Chemical Data Management

Chemical data management is an important process to a number of industries, especially those engaged in manufacturing or research and development. Unfortunately, chemical data is as unwieldy to manage as it is important. This is for a variety of reasons, but the biggest contributing factor is the sheer amount of data available to the public and managed by an enterprise. For decades, chemical data has been recorded in paper, and then excel sheets, and now databases. While efforts have been made to homogenize, or at least centralize this data, there is not one single solution for indexing and searching the wide variety of data types, and sources, managed by an individual enterprise. Chemchart Enterprise changes this paradigm by combining the flexibility of big data with the precision of machine learning, providing a single solution for managing the entire organization chemical space.

ToxTrack Inc, a developer of cheminformatics software, just released its new tool Chemchart Enterprise -- a machine learning (ML) based platform for managing chemical data. Chemchart Enterprise provides a single solution for managing the entire organizational chemical space, combining database management, document processing, and chemical exploration into an intuitive interface.

Built for expansion, Chemchart Enterprise’s internal database can extend to new types of data, while facilitating semantic and structural queries, and providing tools to compare collections of chemicals. Chemchart Enterprise also has the ability to identify and extract chemical references in documents. The platform scans internal document collections (e.g. PDFs, spreadsheets, Word Docs, SDFs) and automatically indexes chemicals into the central database. In addition to organizing internal document repositories, Chemchart Enterprise can join internal data with external datasets like patents, regulatory, or news repositories.

A useful feature in Chemchart Enterprise is being able to get alerts whenever a chemical in the customer's supply chain lands on a new regulatory document, or finds a new use in a patent.

Data management options

The core of Chemchart Enterprise is a chemical database indexed by structure. Each structure is tied to a variety of meta-data, including identifiers, physicochemical properties, and toxicity data. The database can also be expanded to include new sorts of data, whether it be the results of an assay or patent numbers associated with that chemical.

Indexing by structure gives Chemchart Enterprise the foundation for substantial flexibility. For Chemchart Enterprise, flexibility means predictive technologies. Predictive models can be incorporated into this software, using the underlying database as training data to predict for data gaps.

Recognizing that research is the collective work of many researchers, Chemchart Enterprise is a multi-user application with both easily approachable and technical user interfaces. Non-technical users can access an innovative, spreadsheet-like web app to view and contribute to the database. This makes accessing and updating data as simple as using Excel. For users with more technical applications, like building machine learning models, there are standardized application programming interfaces (APIs) with read/write access to the database.