Katana Graph’s AI team exploits graph machine learning-based techniques to lead in accuracy and reduced error rates for drug discovery and molecular activity
AUSTIN, TX - Katana Graph, the AI-powered Graph Intelligence Platform providing faster, deeper, and more accurate insights on massive and complex data, has developed a graph machine-learning based technique to solve twenty-two problems in the Therapeutics Data Commons’ (TDC) molecular property predictions competition, ranking first in eight problems and second in two other problems posed by the competition, among all global participants.
Katana Graph's AI team built a very innovative graph AI model to tackle some of the most challenging problems in the competition, including data-powered drug discovery, property prediction, and learning methods.
The competition is organized by Therapeutic Data Commons, an open-science initiative with AI/ML-ready datasets and AI/ML tasks for therapeutics, spanning the discovery and development of safe and effective medicines. The competition brings together top companies and talent in the AI/ML fields to help pharmaceutical data scientists apply algorithmic techniques to find biomedical and clinical solutions.
The datasets are provided by TDC, and results are submitted to leaderboards for evaluation across a set of factors – including a model’s repeatability, reproducibility, and reusability. The TDC also provides public benchmarks with performance metrics, and the results of the competition have wide-ranging implications for advances in medicine, including reduced timeframes for drug discovery, better-targeted experimentation, and reduced manual effort.
“Katana Graph's AI team built a very innovative graph AI model to tackle some of the most challenging problems in the competition, including data-powered drug discovery, property prediction, and learning methods," said Keshav Pingali, Katana Graph CEO. "I am proud of Katana Graph's stellar performance in TDC's competition, demonstrating not only the far-reaching impact of our graph technology but also the talent and capabilities of our team.”
“Katana Graph’s results, based on their graph ML technique developed on the latest generation Intel® Xeon® Scalable platform, contribute to advancing efforts by the life sciences industry to improve medical treatment and ultimately save lives,” said Greg Lavender, Chief Technology Officer of Intel. “We commend their demonstrated achievements and look forward to collaborating with them to support novel therapeutic approaches that leverage Katana’s AI graph technology to improve the lives of people.”
Katana Graph’s performance in this competition follows its Series A funding announcement in February 2021, and recent partnership with Intel to create and launch a high-performance graph analytics Python library for the benefit of data scientists and the growth of the open core community.