International School of Medicine, Dept. of Biostatistics and Medical Informatics
Istanbul Medipol University, Istanbul, Türkiye

https://sabita.medipol.edu.tr/index.php/portfolio-item/assist-prof-kivanc-kok/

Integrated omics profiling in health and disease

Abstract:
The emergence of omics technologies has enabled high-throughput profiling, screening, and monitoring of biological states, processes, and environments under various conditions at an unprecedented pace. Such innovations have revolutionized the biomedical field and led to the exponential accumulation of vast amounts of biological data. The first omic, genomics, was soon followed by an increasing number of others, such as transcriptomics, proteomics, miRNomics, and metabolomics. Here, the advent of next-generation sequencing (NGS) along with the improvements in mass spectrometry advanced the field of omics and, together with the development of adequate computational solutions and infrastructures, transformed it into a “big data” domain. Interestingly, the human microbiome, recognized as the “second genome,” has added a critical dimension and opened the way for omics-based discoveries at another level of biological complexity, such as host–microbiome interactions and the gut–brain axis. Despite the diversity of current omic platforms and efforts, all omic studies share a common feature that can be divided into two steps: (i) sample preparation and execution (“wet lab”), resulting in raw data generation, and (ii) data analysis (“dry lab”). Bioinformatics has been an integral component and driving force of such data analyses from the very beginning. Remarkably, recent advances in the omics field have enabled greater accuracy (e.g., through deep sequencing and deep proteome profiling) and higher resolution (e.g., through single-cell sequencing and cellular proteomics profiling). The progress also contributed to a more comprehensive (“systems-level”) understanding in terms of integration (e.g., multi-omics analysis) and dynamics (e.g., spatiotemporal omics analysis).An increasing number of omics studies now adapt and leverage machine learning (ML)–based bioinformatics methods for data integration and joint analysis. Harnessing the power of ML has proven especially valuable for handling high-dimensional biological data, uncovering hidden patterns and facilitating the discovery of novel biomarkers. In this regard, both supervised and unsupervised ML techniques play instrumental roles in this rapidly evolving field. Powered by artificial neural networks, deep learning is a driving force reshaping and accelerating omics research. For instance, explainable artificial intelligence holds great promise for advancing personalized medicine, underscoring the need for continued efforts in this direction. Overall, ML-assisted integrated omics profiling offers critical insights into the complex mechanisms underlying health and disease.