MOLECULAR BIOINFORMATICS FOR DISEASE

摘要

This session focuses on the emerging field of molecular bioinformatics techniques applied to understand disease, and establish better prognosis or treatments. With the advent of whole genome association studies, there is a clear need to amass and think about and relate phenotypes in a computational way. For example, Brunner and van Driel showed how syndromes sharing clinical features could be related to the same protein complex (Nature Reviews Genetics 2004 5:545). While the application of methods from the domain of systems biology to the phenotypes of model organisms has enabled the prediction of molecular pathways and gene functions, it is not yet obvious how such methods will scale to phenotypes and diseases of higher organisms. Investigating methods to improve either genetics or phenomics methods independently carries enormous scientific and societal merit in this quest for disease genes. Yet taken together, powerful joint methods are likely to emerge such as phenome-anchored whole genome association studies, where prior probabilities stemming from independent studies can be used to increase, ab initio, the sensitivity of association studies or interpret the associations a posteriori through meta-analyses.