Graph4Med

Graph4Med transforms relational medical databases into graph representations to enable graph-based cohort analysis, mutation-based similarity searches, and visualization of clinical and genomic data including Next Generation Sequencing (NGS) and chromosome microarray results.

Key Features:

• Graph Database Transformation: Converts relational medical databases into a graph data model using Neo4j to integrate heterogeneous data types including NGS and chromosome microarray results.
• Cohort-level Visualization and Analysis: Produces visualizations and cohort analyses of patient attribute distributions such as gender, age, mutations/fusions, and diagnoses to support pattern identification.
• Mutation-Based Similarity Search: Performs mutation- and fusion-based similarity searches to identify patient similarity networks relevant to conditions such as pediatric Acute Lymphoblastic Leukemia (ALL).
• Patient Graph Generation: Generates patient-specific graphs that reveal relationships and patterns within cohorts.

Scientific Applications:

• Pediatric Acute Lymphoblastic Leukemia (ALL) analysis: Applied to a pediatric ALL dataset combining routine health records with NGS and chromosome microarray results to identify patient similarities and mutation-driven subgroupings.

Methodology:

Transforms relational database schemas into a graph data schema and stores the resulting graph in Neo4j.