The Human Genome Project promises to transform genomics into translational discoveries for human health. About 98% of the human genome is non-coding, and its function is gradually being understood in the context of human aging and disease risk. Understanding ‘how’ this non-coding genome contributes to risk in polygenic psychiatric, neurodegenerative, and cancer disorders needs complex computational analyses between available human genetics and related clinical biomarkers in disease. This will help build a powerful algorithm to predict disease risk in healthy subjects and aid in preventative intervention in healthy individuals from developing a disease.
UKBiobank is a world leader in collecting and cataloging accurate clinical and phenotypical datasets from diverse human subjects, providing an ideal platform for genetic risk outcome prediction analyses. BioTuring excels in developing ML frameworks and analyses and has curated several single cell type datasets that can be tested in the context of disease risk. Not all human cells develop disease therefore identifying the correct cell-type populations ‘at risk’ of contributing to disease in specific individual (genetic predictor tests) will help focus preventative diagnosis to keep these people healthy longer.
With the help of UKBiobank resources, BioTuring scientists will develop a ML framework integrating a range of phenotypical measures associated with the genetic sequence. This framework would aim to predict disease risk profiles based on the 98% non-coding genetic sequences, offering potential for polygenic psychiatric, neurodegenerative, and cancer diseases and other diseases as well. The ML framework will integrate existing and future publicly available datasets to refine single cell-type phenotypical (disease risk) predictions. Additionally, our study will enhance therapeutic candidate identification from genetic risk associated therapeutic pathways in many diseases, where clinical discovery and intervention are significant needs. This endeavor represents one of the initial comprehensive efforts to decipher cell-type polygenic risk translations published across several disease areas spanning mental disorders and cancer in the last two decades.
