Utilizing Omic Data to Understand Integrative Physiology

Over the past several decades, physiological research has undergone a progressive shift toward greater and greater reductionism, culminating in the rise of "molecular physiology." The introduction of omic techniques, chiefly protein mass spectrometry and next-generation DNA sequencing (NGS), has further accelerated this trend, adding massive amounts of information about individual genes, mRNA transcripts, and proteins. However, the long-term goal of understanding physiological and pathophysiological processes at a whole organism level has not been fully realized. This review summarizes the major protein mass spectrometry and NGS techniques relevant to physiology and explores the challenges of merging data from omic methodologies with data from traditional hypothesis-driven research to broaden the understanding of physiological mechanisms. It summarizes recent progress in large-scale data integration through 1) creation of online user-friendly omic data resources with cross-indexing across datasets to democratize access to omic data; 2) application of Bayesian methods to combine data from multiple omic datasets with knowledge from hypothesis-driven studies to address specific physiological and pathophysiological questions; and 3) application of concepts from natural language processing to probe the literature and to create user-friendly causal graphs representing physiological mechanisms. Progress in development of so-called large language models, e.g. ChatGPT, for knowledge integration is also described, along with a discussion of the shortcomings of large language models with regard to management and integration of physiological data.