Text Mining and Data Modeling of Karyotypes to aid in Drug Repurposing Efforts

Karyotyping, or visually examining and recording chromosomal abnormalities, is commonly used to diagnose and treat disease. Karyotypes are written in the International System for Human Cytogenetic Nomenclature (ISCN), a computationally non-readable language that precludes full analysis of these genomic data. In response, we developed a cytogenetic platform that transfers the ISCN karyotypes to a machine-readable model available for computational analysis. Here we use cytogenetic data from the National Cancer Institute (NCI)-curated Mitelman database1 to create a structured karyotype language. Then, drug-gene-disease triplets are generated via a computational pipeline connecting public drug-gene interaction data sources to identify potential drug repurposing opportunities.