AUTHOR=Wang Yingying , Wang Lili , Liu Yinhe , Li Keshen , Zhao Honglei 

TITLE=Network Analyses Based on Machine Learning Methods to Quantify Effects of Peptide–Protein Complexes as Drug Targets Using Cinnamon in Cardiovascular Diseases and Metabolic Syndrome as a Case Study

JOURNAL=Frontiers in Genetics

VOLUME=Volume 12 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2021.816131

DOI=10.3389/fgene.2021.816131

ISSN=1664-8021

ABSTRACT=Peptide-protein complexes play important roles in multiple diseases such as cardiovascular diseases (CVDs) and metabolic syndrome (MetS). The peptides may be the key molecules in the designing of inhibitors or drug targets. Many Chinese traditional drugs are shown to play various roles in different diseases and the comprehensive analyses should be performed using networks which could offer more information than results generated from a single level. In this study, we designed a network analysis pipeline based on machine learning methods to quantify the effects of peptide-protein complexes as drug targets. Three steps including pathway filter, combined network construction, biomarker prediction and validation based on peptides were performed using cinnamon (CA) in CVDs and MetS as a case. Results showed that 17 peptide-protein complexes including 6 peptides and 4 proteins were identified as CA targets. The expressions of AKT1, AKT2, and ENOS were tested using qRT-PCR in a mouse model we constructed. AKT2 was shown to be CA indicating biomarker while E2F1 and ENOS were CA treatment targets. AKT1 was considered as diabetic responsive biomarker since it was down-regulated in diabetic but not related to CA. Taken together, our pipeline could identify new drug targets based on biological function analyses. This may provide a deep understanding of the drugs’ roles in different diseases which may foster the development of peptide-protein complexes-based therapeutic approaches.