Integration of solutions and services for multi-omics data analysis towards personalized medicine

Current advances in high-throughput and imaging technologies are paving the way next-generation healthcare, tailored to the clinical and molecular characteristics of each patient. The Big Data obtained from these technologies are of little value to society unless it can be analyzed, interpreted, and applied in a relatively customized and inexpensive way.We propose a flexible decision support system called IntelliOmics for multi-omics data analysis constituted with well-designed and maintained components with open license for both personal and commercial use. Our proposition aims to serve some insight how to build your own local end-to-end service towards personalized medicine: from raw data upload, intelligent integration and exploration to detailed analysis accompanying clinical medical reports. The high-throughput data is effectively collected and processed in a parallel and distributed manner using the Hadoop framework and user-defined scripts. Heterogeneous data transformation performed mainly on the Apache Hive is then integrated into a so called ‘knowledge base’. On its basis, manual analysis in the form of hierarchical rules can be performed as well as automatic data analysis with Apache Spark and machine learning library MLlib. Finally, diagnostic and prognostic tools, charts, tables, statistical tests and print-ready clinical reports for an individual or group of patients are provided. The experimental evaluation was performed as part of the clinical decision support for targeted therapy in non-small cell lung cancer. The system managed to successfully process over a hundred of multi-omic patient data and offers various functionalities for different types of users: researchers, bio-statisticians/bioinformaticians, clinicians and medical board.