Apply support vector regression to extract the potential susceptibility genes of chronic obstructive pulmonary disease

Abstract

Chronic obstructive pulmonary disease (COPD) is a complex disorder classified as the 3rd cause of the death worldwide. So far, we know that this disease is progressive and can not be cured. In recent years, although some genes have been reported to be associated with COPD, the overlapped genetic associations can't be replicated. Therefore, it is difficult to synthesize and interpret these different findings. To address this issue, we conducted an integrated data analysis by combining network topological properties with support vector regression (SVR) to extract the potential susceptibility genes of COPD. As a result, COPD-related risk genes such as BBS9, ADAM19 and TGFB1 were identified, and these genes were supported by some previous and recent evidences. Our approach can help improve the accuracy in identifying COPD-related risk genes.