Solving the Biological Complexity with Data-Driven Machine Learning Methods


KHAS Core Program

December 28, Tuesday, 20:00

Biology and its tools have been used since ancient times to understand nature and find cures to diseases, and from then, its applications evolved with an extraordinary speed. Biological interactions in organisms are complex and require interdisciplinary complex systems understanding. This complexity begins with the number of protein-encoding genes found in human cells. Each cell in the body shares the same genes, but their combinatorial expression levels (transcripts and proteins) result in diverse cell- and tissue types. Therefore, biological complexity increases in the transcript and protein levels. Considering that the Human Genome Project took 13 years to complete the sequencing of human reference genome data, understanding the human transcriptome and proteome map is in its very early days.

This talk focuses on a recently developed state-of-the-art technology field called single-cell omics. Its application areas, together with its limitations, will be explained. These limitations can be elucidated with interdisciplinary approaches. Our newly developed machine learning method is successfully applied to solve one of the current problems in the single-cell omics field. This talk will cover how one significant feature of nature was used to decipher this current problem. Finally, the importance of this analysis for integrating high-dimensional biological data and the relevance of creating cellular maps in organisms will be clarified.