Streamlining multi-omic and artificial intelligence analysis through interrogative biology and baicis for translational precision medicine applications in clinical oncology

Abstract

Background: Despite advances in high throughput molecular technologies, increased availability of clinical information, and access to complex population level datasets, translating this information into causal and actionable clinical guidance in oncology remains a challenge. Method(s): The BERG Interrogative Biology platform deconstructs the established paradigm by using patient biology to guide the entire drug development process from R&D to clinic, leading to improved clinical outcome. In order to properly characterize the molecular phenotype of patients or disease states, this platform allows for systematic interrogation of each biological sample by high-throughput multi-omic technologies such as proteomics, lipidomics and metabolomics. This is then combined with further analytical methods that allows for assessment of sample quality through statistical, environmental/demographic influence, sample handling, and pharmacological impact markers to elucidate causal molecular signal from inherent noise. Result(s): BERG ETL System uses a proprietary data-driven algorithm to automatically extract, normalize, correct eventual systematic errors, align and unify all data sources and types, outputting a harmonized molecular and/or clinical profile, which can be used for summary reports such as patient dashboards, standard analysis such as statistics and machine learning, and to be analyzed by BERG's Artificial Intelligence (AI) Technology, bAIcis. When applied to clinical trial information, bAIcis uses a multilayer method to identify clinical and molecular markers that can stratify patients based on trial outcomes such as ?Response to Treatment?, ?Quality of Life? or ?Adverse Events? as well as identification of disease drivers. Conclusion(s): Using this approach a comprehensive understanding of causal drivers, predictive biomarkers aligned with therapeutic benefit, and identification of adverse event populations in cancer indications can be elucidated. This is streamlined through an AI driven platform based on quality metric to support precision medicine in oncology drug development.