Highly Scalable A.I. Analytics for Life Sciences R&D
SDF File Handling for Compound Structure Analytics
Advanced A.I. Chemistry Analytics
Highly Scalable Image Management & A.I. Analytics
ChemVector Enables de novo Compound Structure Generation
Target Finding & Drug Repurposing
Vyasa’s Neural Concept Recognition A.I. technology can train on concept types (e.g. drugs, diseases, pathways, conditions, side effects, genes) in structured and unstructured content. Once trained Cortex builds a dynamic knowledge graph of everything known about those concepts across all sources. Because everything in Cortex is converted into a vector space, each concept in the knowledge graph is connected to everything else in vector space. This allows users to query Cortex for proximity of concepts across very large knowledge sets thereby yielding unexpected relationships between mechanisms of action and disease conditions that might be otherwise missed. This type of vector space analysis is an excellent tool for drug repurposing and target finding.
A.I. Driven de novo Compound Design
Vyasa has developed a proprietary deep learning algorithm for /de novo/ small compound design called ChemVector. ChemVector utilizes an autoencoder-based neural network that can achieve >98% reconstruction accuracy on SMILES strings along with Bayesian optimization to identify and generate novel compounds that optimize variables such as log-p, molecular weight and synthetic viability. ChemVector is an analytic module available in Vyasa Cortex.
Deep Learning Chemistry Analytical Library
Cortex provides a wide range of deep learning analytical modules for chemical analysis. From toxicity analysis to molecular modeling, Vyasa Cortex makes it easy to apply deep learning algorithms to compounds sets imported into Cortex.
Electronic Laboratory Notebook Mining
Electronic Laboratory Notebooks (ELN’s) are widely used tools for capturing the research performed in laboratories. It is sometimes challenging though to gain a bird’s eye view of the information stored in these systems and the potential connections identified in the content. Vyasa’s Neural Concept Recognition technology can be trained to scan for specific types of content in ELN records (e.g. compound structures, drug names, reaction mechanisms, reagents, instrumentation) and enable a user to conduct analytics on the concepts found. Vyasa Cortex provides organizations with more effective discovery tools for the valuable content stored in their ELN systems.