HiberCell and Strasbourg Institute for Precision Medicine Publish Machine Learning Data and Approach on Causal Biology of Severe COVID-19

Artificial intelligence, machine learning (AI/ML) approach, developed in cancer and cardiovascular disease, identifies gene expression network in acute respiratory distress syndrome (ARDS) in a subset of COVID-19 patients published in Science Translational Medicine

HiberCell, a clinical stage biotechnology company developing therapeutics to treat cancer relapse and metastasis, and its academic collaborators in Strasbourg (France) published data from a study focused on defining biological networks believed responsible for severe and often deadly acute respiratory distress syndrome (ARDS) in a younger COVID-19 patient cohort. This study, entitled the “Identification of driver genes for severe forms of COVID-19 in a deeply phenotyped younger patient cohort,” appears online in Science Translational Medicine.

Focused on the underlying biology of COVID-19 in the absence of confounding factors that include age and underlying health issues, the study studied a subgroup of COVID-19 patients that presented with no underlying medical condition and were younger than 50 years of age. The aim of the study was to define biological networks responsible for severe and often deadly ARDS in COVID-19 patients. Mapping and ranking the putative interactions between a network of 600 genes in these patients showed that abnormal ADAM9 expression was a key causal driver of COVID-induced ARDS.

“The study further supports the utility of our biologically-validated AI/ML platform that is focused on advancing our portfolio of cancer therapeutics,” said study co-principal investigator, Tom Chittenden, PhD, DPhil, PStat, chief technology officer of HiberCell and the co-founder of HiberCell’s AI initiative. “This study underscores the disease-agnostic potential of our causal biology, AI/ML approach that is well positioned to provide disease insights even with limited patient cohorts.”

“This is an important example of the potential impact machine learning brings to life sciences,” commented Daniel Lidar, PhD, Viterbi Professor of Engineering at the University of Southern California (USC) and a HiberCell collaborator. “Using multiple independent ML methods, including a quantum variant, this study builds upon our work in breast cancer, further validating the utility of unconventional ‘Ising-type’ ML in quantum computing. It builds upon results published in April in which we identified replicable signals in high-dimensionality data in small cohorts.”

“This work provides a deeper understanding of COVID-19 in a younger, healthier patient subset,” said co-principal investigator, Seiamak Bahram, MD, PhD, chair of immunology at the University of Strasbourg Faculty of Medicine, INSERM Unit 1109 and Strasbourg University Hospitals.  “We have further supported these findings in additional cohorts, solidifying the identification of these biological networks. We are actively working to further elucidate the importance of ADAM9 in virus uptake and replication in human lung cells. These data may provide an alternative therapeutic approach for COVID-19 or in broader terms, ARDS in general.”

“Recognizing the need for additional research, these data highlight the potential of our patient-centric AI/ML initiative at HiberCell that is driving pharmacodynamic biomarker validation across our first-in-class pipeline of stress modulators in cancer,” said Alan Rigby, President and CEO of HiberCell.