OpenOmics: A bioinformatics API to integrate multi-omics datasets and interface with public databases.

Leveraging large-scale multi-omics data is emerging as the primary approach for systemic research of human diseases and general biological processes. As data integration and feature engineering are the vital steps in these bioinformatics projects, there currently lacks a tool for standardized preprocessing of heterogeneous multi-omics and annotation data within the context of a clinical cohort. OpenOmics is a Python library for integrating heterogeneous multi-omics data and interfacing with popular public annotation databases, e.g., GENCODE, Ensembl, BioGRID. The library is designed to be highly flexible to allow the user to parameterize the construction of integrated datasets, interactive to assist complex data exploratory analyses, and scalable to facilitate working with large datasets on standard machines. In this paper, we demonstrate the software design choices to support the wide-ranging use cases of OpenOmics with the goal of maximizing usability and reproducibility of the data integration framework.