Machine Learning applied to multiomics analysis for exploration of microbial communities

Abstract

As high-throughput sequencing technologies have emerged in recent decades, the rate of biological data collection has been amplified, at the same time as this has served as a bottleneck for further processing and analysis, given the sheer volume of this data. Considering the voluminous data coming from these technologies, the creation of fields of studies encompassing the so-called Omics Sciences, which involves fields such as genomics, transcriptomics, proteomics, and metabolomics, has been enabled. With the availability of these data, the need for integration of these different fields has been observed, to create a multifaceted analysis, making possible the discovery of, for example, new biomarkers, which may have the potential to help in disease prediction and contribute to the development of personalized medicine. Given this essential integration, this project aims to explore Machine Learning methods, such as those available in the BioAutoML tool, to integrate omics data and generate a multi-omics pipeline with a focus on complex microbial communities.