Niven Narain, is the CEO and President of BERG Health,a clinical-stage, AI-powered biotechnology company taking a bold ‘Back to Biology™’ approach to healthcare. By leveraging a proprietary intelligence platform – Interrogative Biology®, the goal is to map disease and revolutionize treatments for patients around the world.
You are one of the Co-Founders of BERG Health. What inspired launching a startup that fused AI with biotech?
Truthfully, my inspiration and vision came from frustration of the methodical, predictable, and long, costly process of drug development taking 12-14 years and over $2.6BN to advance a drug treatment to market. Additionally, Berg was keen on taking the biases out of the process and wanted to define a path that would be patient-centric. Our approach blends a patient’s own data, which has been deeply screened for the full layers of biology, with a back-ended Bayesian AI system to funnel all the patient data, and in turn, drive hypotheses.
What are some of the different diseases or cancers that are being targeted?
We are currently targeting diseases across oncology, neurology and rare diseases including:
- Immunology/Inflammatory Diseases (LUPUS)
- Cardiovascular & Metabolic Disease (Diabetes, NASH/NAFLD)
- Neurological Diseases (Parkinson’s Disease, Alzheimer’s Disease & Autism Spectrum Disorders)
- Rare Disease (Epidermolysis Bullosa)
- Multiple Cancers (Glioblastoma, Pancreatic, Breast, Prostate, and other highly aggressive cancers)
BERG Health uses its proprietary platform Interrogative Biology® to map out diseases and treatment options. Could you elaborate on what precisely Interrogative Biology® is?
The platform encompasses a process in which a disease-specific model is built from human-derived biospecimens (making it relevant to human disease rather than surrogate animal models). The disease model is subjected to comprehensive molecular profiling (multi-omic defragmentation – genomics, transcriptomics, proteomics, lipidomics, metabolomics) to generate trillions of molecular data points. This data is integrated to individual patient information (clinical and real world information) using BERG’s proprietary Bayesian Artificial Intelligence (AI) algorithm(s) to generate molecular maps of disease that are compared to non-diseased controls to identify novel biology underlying disease. The output is, then, subjected to stringent wet-lab validation techniques for disease-relevant functional models and CRISPR.
Can you explain how BERG Health crowdsources innovation?
BERG collaborates with academic and clinical/medical institutions to generate biobanks of high-quality, clinically-annotated biospecimens for disease-specific programs. BERG has built a team of in-house experts who specialize in some of the multi-omic capabilities. The team also collaborates with groups with specific expertise for additional data profiles, e.g. genomic and transcriptomic profiles which are generated with collaborators. We collaborate with leading clinical/medical institutions and state/federal agencies, including the U.S. Department of Defense and Department of Energy-funded Oak Ridge National Lab (ORNL), among others. We have seen firsthand the importance of collaboration and its role in generating high-fidelity data backed by patient medical histories and real-world data, an essential step for integration to the biological models. Academic collaborations are vital for wet-lab validation of in-silico outputs to generate a novel, scientific foundation of specific molecular insights delineated by the platform. Berg also engages the insights and feedback from KOL’s at the commencement of building AI models, and uses these collaborations to do independent validation of the platform output.
Can you describe how AI is being used to discover elements that are triggering diseases or cancers?
BERG’s Bayesian AI uses internally generated/curated data for de novo discovery of disease specific triggers – identify targets for treatment, biomarkers for diagnosis, staging/stratification, companion diagnostics for response/outcome and longitudinal molecular mapping for generating fingerprints of response and adverse events. The AI compares models of disease vs. non-disease populations and the “delta” network infers the trigger points of disease.
What are some ways in which AI assists with locating biomarkers in patients that are failing to respond to certain therapies or vaccines?
By capturing the full narrative of patient biology through multi-omic analysis and employing Bayesian causal inference from longitudinal patient samples, BERG is able to identify causal signals of response to known therapies, repurposing opportunities and signatures of molecular entities impacting viral titer and durability of response of vaccines. Multi-omics goes well beyond the genome and allows you to identify circulating factors that impact health outcomes.
What are some of the current drugs in your pipeline?
BERG has several programs in clinical and preclinical development with the most mature assets in oncology & neurological disease.
- In Oncology, BPM 31510-IV is a novel small molecule targeting cancer cell metabolism that has successfully completed a Phase 1 (safety/tolerability) in solid tumors and brain cancer (GBM). BPM 31510-IV Phase 2 (efficacy, pancreatic cancer). This is currently in clinical development for GBM (Ph2/3) and Ph3 Pancreatic Cancer.
- BPM 31510-Topical – Successful completion of Ph1 in Epidermolysis Bullosa (Rare Disease/Orphan Designation), planning Ph2/3 clinical development.
- BPM 31510-Oral – Successful completion of Ph1 oral in health human volunteers, in early planning stages of Ph2 clinical development for oncology/non-oncology indications.
- BPM 31543 is a small molecule for prevention of chemotherapy induced alopecia. Safety & Tolerability with initial signal of efficacy has been established in Phase 1 clinical trial. This asset is currently in planning for Ph2/3 clinical development.
- BPM 42522 – 1st in class small molecule targeting a novel BERG platform identified target in the Ubiquitin Proteosome Pathway in IND enabling studies for FIH studies in oncology Q1FY2021.
- A novel target identified for Parkinson’s Disease is currently in drug discovery efforts.
BERG Health has recently partnered with the Department of Energy’s Oak Ridge National Lab. Could you give us some details on what this partnership entails?
BERG’s platform has the capacity to generate list of several potential targets for disease intervention. The Summit supercomputer at ORNL has the capabilities for analyzing comprehensive molecular structures of these targets and identifying small drug-like molecules that can be utilized for rapid validation of the targets, leading to shortened target validation/drug discovery development times for new therapeutics. The power of Summit’s computational capabilities reduce drug discovery process from several months/years to a few hours/days to generate initial high-quality “hits” of drug-like molecules. The BERG-ORNL collaboration provides the foundation of rapid identification of novel disease-specific targets and associated drug discovery process. It further enables seamless generation of disease-specific product pipelines primed for clinical development. The major impact of this collaboration is on the time/cost of discovery and development of new drugs.
COVID-19 is obviously on everyone’s minds, how is BERG Health assisting with this effort?
BERG’s AI-enabled Interrogative Biology Platform has been leveraged to generate a COVID-19 specific model, resulting in the identification of several known and novel targets with potential to impact the time course of infection and potential repurposing of approved drugs to minimize/mitigate clinical outcomes. Through our active collaboration with ORNL, we are currently engaged in the discovery and development of small molecules against novel targets for potential treatment of COVID-19.
Thank you for the interview, readers who wish to learn more should visit BERG Health.