Detection of Disease Clusters and Functional Characterization

PROJECT TITLE :

Disease Cluster Detection and Functional Characterization

ABSTRACT:

With the advancement of molecular biology, the mechanisms behind human diseases have been identified. The understanding of the underlying molecular basis of illnesses is important for disease prevention, diagnosis, and treatment. Positional cloning of disease genes and genome-wide association analyses have been used to investigate disease-gene associations for decades. Many clinical cases, in particular, have identified connections between various disorders. The shared disease-associated genes could help show the inherent relationship at the genetic level, provide a way to compare diseases, and build a human disease network. Although numerous methods for measuring illness similarity have been presented, they only evaluate genes or functions that are directly linked to diseases and overlook interactions between genes or functions. These interactions lead to a lack of disease classification accuracy. We proposed a systematic investigation based on a network-based disease module to further investigate the relationship between different human diseases and determine whether this correlation is dependent on the actions of corresponding disease genes. On the one hand, 299 diseases are divided into 15 relatively separated disease clusters using a disease clustering method based on disease separation scores. On the other hand, based on disease-associated genes, their GO keywords, and KEGG pathway annotations, an optimal clustering scheme differentiating 15 disease groupings was learned. The main signatures discovered were the most useful in differentiating different illness clusters and reflected critical processes in pathogenesis. This research presents a novel method for predicting network and function features, as well as revealing the functional essence of disorders.

Did you like this research project?

To get this research project Guidelines, Training and Code... Click Here