Multi-factor regulatory network and different clusters in hypertrophic obstructive cardiomyopathy

BMC Medical Genomics

Lei Huang¹  Jian Zhuang²  Xianyu Qin³  Shaoxian Chen⁴  Pengju Wen⁴  Yueheng Wu⁴  Sicheng Chen⁵ 

[1] Department of Cardiology, Guangdong Cardiovascular Institute, Guangdong Provincial People’s Hospital, Guangdong Academy of Medical Sciences, Guangdong, China;Department of Cardiovascular Surgery, Guangdong Cardiovascular Institute, Guangdong Provincial Key Laboratory of South China Structural Heart Disease, Guangdong Provincial People’s Hospital, Guangdong Academy of Medical Sciences, Guangdong, China;Department of Thoracic Surgery, Thoracic Cancer Center, The Sixth Affiliated Hospital of Sun Yat-Sen University, Guangzhou, China;Department of Cardiovascular Surgery, Guangdong Cardiovascular Institute, Guangdong Provincial Key Laboratory of South China Structural Heart Disease, Guangdong Provincial People’s Hospital, Guangdong Academy of Medical Sciences, Guangdong, China;Guangdong Cardiovascular Institute, Guangdong Provincial Key Laboratory of South China Structural Heart Disease, Guangdong Provincial People’s Hospital, Guangdong Academy of Medical Sciences, Guangdong, China;Shantou University Medical College, Shantou, China;Department of Cardiology, Guangdong Cardiovascular Institute, Guangdong Provincial People’s Hospital, Guangdong Academy of Medical Sciences, Guangdong, China;

关键词: Hypertrophic obstructive cardiomyopathy;    Weighted gene co-expression network analysis;    Multi-factor regulatory network;    Different clusters;   

DOI  :  10.1186/s12920-021-01036-4

来源: Springer

【 摘 要 】

BackgroundPractical biosignatures and thorough understanding of regulatory processes of hypertrophic obstructive cardiomyopathy (HOCM) are still lacking.MethodsFirstly, public data from GSE36961 and GSE89714 datasets of Gene Expression Omnibus (GEO), Gene database of NCBI (National Center of Biotechnology Information) and Online Mendelian Inheritance in Man (OMIM) database were merged into a candidate gene set of HOCM. Secondly, weighted gene co-expression network analysis (WGCNA) for the candidate gene set was carried out to determine premier co-expressed genes. Thirdly, significant regulators were found out by virtue of a multi-factor regulatory network of long non-coding RNAs (lncRNAs), messenger RNAs (mRNAs), microRNAs (miRNAs) and transcription factors (TFs) with molecule interreactions from starBase v2.0 database and TRRUST v2 database. Ultimately, HOCM unsupervised clustering and “tsne” dimensionality reduction was employed to gain hub genes, whose classification performance was evaluated by a multinomial model of lasso logistic regression analysis binded with receiver operating characteristic (ROC) curve.ResultsTwo HOCM remarkably-interrelated modules were from WGCNA, followed by the recognition of 32 crucial co-expressed genes. The multi-factor regulatory network disclosed 7 primary regulatory agents, containing lncRNAs (XIST, MALAT1, and H19), TFs (SPI1 and SP1) and miRNAs (hsa-miR-29b-39 and has-miR-29a-3p). Four clusters of HOCM and 4 hub genes (COMP, FMOD, AEBP1 and SULF1) significantly expressing in preceding four subtypes were obtained, while ROC curve demonstrated satisfactory performance of clustering and 4 genes.ConclusionsOur consequences furnish valuable resource which may bring about prospective mechanistic and therapeutic anatomization in HOCM.

【 授权许可】

CC BY