Exploring Integrative Analysis using the BioMedical Evidence Graph

Abstract

Highlights• Data resource connected extremely diverse set of cancer data sets • Graph query engine that can be easily deployed and used on new datasets • Easily installed python client • Server online at bmeg.io AbstractThe analysis of cancer biology data involves extremely heterogeneous datasets including information from RNA sequencing, genome-wide copy number, DNA methylation data reporting on epigenomic regulation, somatic mutations from whole-exome or whole-genome analyses, pathology estimates from imaging sections or subtyping, drug response or other treatment outcomes, and various other clinical and phenotypic measurements. Bringing these different resources into a common framework, with a data model that allows for complex relationships as well as dense vectors of features, will unlock integrative analysis. We introduce a graph database and query engine for discovery and analysis of cancer biology, called the BioMedical Evidence Graph (BMEG). The BMEG is unique from other biological data graphs in that sample level molecular information is connected to reference knowledge bases. It combines gene expression and mutation data, with drug response experiments, pathway information databases and literature derived associations. The construction of the BMEG has resulted in a graph containing over 36M vertices and 29M edges. The BMEG system provides a graph query based API to enable analysis, with client code available for Python, Javascript and R, and a server online at bmeg.io. Using this system we have developed several forms of integrated analysis to demonstrate the utility of the system. The BMEG is an evolving resource dedicated to enabling integrative analysis. We have demonstrated queries on the system that illustrate mutation significance analysis, drug response machine learning, patient level knowledge base queries and pathway level analysis. We have compared the resulting graph to other available integrated graph systems, and demonstrated that it is unique in the scale of the graph and the type of data it makes available. SummaryThe analysis of cancer biology data involves extremely heterogeneous datasets including information. Bringing these different resources into a common framework, with a data model that allows for complex relationships as well as dense vectors of features, will unlock integrative analysis. We introduce a graph database and query engine for discovery and analysis of cancer biology, called the BioMedical Evidence Graph (BMEG). The construction of the BMEG has resulted in a graph containing over 36M vertices and 29M edges. The BMEG system provides a graph query based API to enable analysis, with client code available for Python, Javascript and R, and a server online at bmeg.io. Using this system we have developed several forms of integrated analysis to demonstrate the utility of the system. Publication:ncbi.nlm.nih.gov/pubmed/27269946 1033 Publication:ncbi.nlm.nih.gov/pubmed/27174596 1029 Publication:ncbi.nlm.nih.gov/pubmed/19223544 858 Publication:ncbi.nlm.nih.gov/pubmed/20619739 6...

Advances in medical knowledge graph construction techniques and applications

“…The content of the knowledge graph to the previous search engine for innovation focuses on the expansion of the query, through the knowledge graph to retrieve the entities involved in the query content and the connectivity between the entities, and also to achieve the provision of more attribute content, so as to meet the user's query needs. Struck Adam and other scholars [11] proposed that medical knowledge graphs have become a common method for representing biomedical knowledge and is to achieve effective combination of information retrieval and UMLS to achieve query expansion, which is then widely applied to the medical field. Some scholars achieved query expansion by incorporating medical ontology MeSH into the search engine, involving entities of synonymy, proximity and other concepts as well as linkage, so that the efficiency of information retrieval can be effectively improved.…”