Beyond Mutation Burden and PD-L1: Deep Molecular and Immune-Infiltrate Stratification of Tumors With the Data4Cure Biomedical Intelligence® Could

Data4Cure, Inc. a biomedical data-to-knowledge company will take part in the 2017 BioData West conference held in San Francisco, April 26-27. Janusz Dutkowski, Ph.D., CEO, will participate in a panel on the future of AI in pharmaceutical development and healthcare and give a talk on the second day of the conference. 

As machine learning and AI are increasingly entering the biomedical domain, it is important to consider how these approaches can extend biomedical knowledge and not add to the complexity and noise that often hamper the progress in the field,” said Dutkowski. “At Data4Cure we believe in building advanced computational tools that are guided by prior knowledge of how cells and biological systems work. This prior knowledge serves to narrow down the search space of possible solutions and guides AI and machine learning algorithms towards biologically-sound solutions that have a greater chance of success in clinical development. This guided approach is at the core of what the Biomedical Intelligence Cloud does,” said Dutkowski. 

The talk on day two of the conference will focus on Data4Cure’s model-based molecular and immune infiltrate stratification engine which identifies cancer subtypes that may benefit from immuno-oncology and targeted therapies. Data4Cure’s solution uniquely combines molecular network and pathway information with somatic mutation and immune-infiltration profiles inferred from RNA sequencing data. This approach looks beyond mutation burden and PD-L1 expression to identify subtypes based on their distinct biology. This includes detailed characterization of key cancer pathways driving each subtype as well as determining the state of the immune system infiltrating the tumor. “Across multiple cancer types from colorectal to lung and kidney cancer we are finding that our systems approach can provide our pharmaceutical partners with a lot more information and is also more predictive than currently used methods based on mutation burden and/or PD-L1 expression,” said Dutkowski.

The stratification and immuno-oncology tools are integrated with CURIE™, a dynamic biomedical knowledge graph that powers the Biomedical Intelligence Cloud. “CURIE is continuously updated with information mined from a variety of data sources. It cross-references tens of thousands of datasets, bioinformatics analyses, literature, clinical trials and external databases, and provides immediate data-driven answers to over 100 million biomedical questions helping to identify novel disease mechanisms, biomarkers and therapeutic targets,” said Roy Ronen, Ph.D., VP, Computational Biology at Data4Cure, who will also be presenting Data4Cure’s platform at the conference.

Launched in 2016, the Biomedical Intelligence Cloud is used across some of the largest global pharmaceutical companies and leading research institutes. “We are very excited to work with some of the best research groups in the industry and academia,” said Dutkowski. “By dramatically accelerating the access to data-driven knowledge and uncovering non-obvious connections hidden in millions of datasets, scientific papers and clinical trials we can really start to turn data into cures.

Source: Data4Cure Inc.