Speaker: Dr. Zhengjun Zhang, Professor of Statistics, University of Wisconsin

Abstract: Genes functionally associated with SARS-CoV-2 and genes functionally related to COVID-19 disease can be different, whose distinction will become the first essential step for successfully fighting against the COVID-19 pandemic. Unfortunately, this first step has not been completed in all biological and medical research. This talk introduces a nearly perfect classifier with the smallest subset and smallest number of signatures (S4) of genes to solve the problem. The S4 classifier is theoretically proved to be efficient with exceptional predicting power. The final classifiers are interpretable with clear signature patterns and functional effects, significantly different from existing ones. The S4 classifiers lead to 100% accuracy in classifying hospitalized patients, including ICU patients, with COVID-19 disease and other non-COVID-19 diseases into their respective groups using five critical genes from their blood sampled gene expressions. The S4 classifiers lead to the best reported 91.88% accuracy for PCR sampled SARS-CoV-2 data with a set of five critical genes. This talk is going to show that genes and their transcriptional response and functional effects to SARS-CoV-2 and genes and their functional signature patterns to COVID-19 antibody are significantly different, which can be interpreted as the former is the point of a phenomenon, and the latter is the essence of the disease. Such significant findings can help explore the causal and pathological clue between SARS-CoV-2 and COVID-19 disease and fight against the disease with more targeted vaccines, antiviral drugs, and therapies. Applying the S4 classifiers to cancer studies again leads to almost 100% accuracy of dozens of trials and thousands of patients. This talk will present results in lung cancer and breast cancer studies. The breast cancer studies show that the widely targeted eight breast cancer-related genes in the literature and medical practice lack predicting power compared with newly identified six genes, which leads to 100% accuracy. If time allows, this talk will also introduce a new model called the Absolute and Relative Treatment Effects (AbRelaTEs) model, which is viewed as a generalization of logistic regression and an enhanced model with increased flexibility and interpretability and applicability in real data applications than the logistic regression.