Benzinga | Jan 13, 2022 11:47AM EST

Predictive Oncology Reports Prelim. Results Of Its Discovery 2021 Program: Its Helomics Unit Showed Proprietary AI Porogram CoRE Could Identify Potential New Chemo Drug Uses

Predictive Oncology Inc. (NASDAQ:POAI) reported the preliminary results of its Discovery 2021 program today, announcing that its Helomics subsidiary's evaluation demonstrated their proprietary Artificial Intelligence (AI) program CoRE(tm) (Computational Research Engine), combined with their tumor profile data and matched ovarian cancer samples, could identify potential new chemotherapy drug uses. These drugs are not currently approved for ovarian cancer, but, with further study, could be used for ovarian cancer treatment following clinical trial and regulatory approval.

The Discovery 2021 program has combined Helomics' proprietary knowledgebase, its AI Machine Learning Program and its TruTumor(tm) platform to create PeDAL(tm) (Patient-centric Discovery by Active Learning). PeDAL is a unique approach to drug discovery, accelerating the selection process to identify potential lead compounds for use in all cancers, not just ovarian. POAI is ready to partner with biopharmaceutical companies.

Active learning-driven experimentation iteratively improves a predictive model until it reaches a specific desired goal. Helomics' evaluation demonstrated that PeDAL made confident predictions for at least 9 times the number of drug-cancer sample combinations studied while identifying potential drugs for repurposing in ovarian cancer treatment. "The study constructed a model of the effects of 175 cancer drugs on 130 patient tumor cell lines. The results indicate that after measuring only 2.6% of all combinations of drugs and cell lines, high confidence predictions could be made for an additional 24% of the combinations," said Dr. Robert F. Murphy, Professor of Computational Biology Emeritus at Carnegie Mellon University and co-inventor of CoRE and a member of Predictive Oncology's Scientific Advisory Board. PeDAL predicted drugs to inhibit cancer in up to 40% of the samples in the study. In addition, variation was observed in the effectiveness of different drugs with the same identified mechanism of action.

"PeDAL leads to a better selection of drugs, matching the different cancer types to the patients associated with those cancers and potential treatments, improving the scientific and clinical success of drug development. This also translates to millions of dollars in reduced development costs for our future partners. PeDAL can now deliver a proven system that efficiently drives drug response experiments and pulls patient-centric data into the discovery process," said J. Melville (Mel) Engle, CEO, Chairman of Predictive Oncology Inc.