Machine learning and bioinformatics to identify 8 autophagy-related biomarkers and construct gene regulatory networks

Fengjun Zhang1, Mingyue Xia2, Jiarong Jiang3

  • 1College of Acupuncture and Massage, Shandong University of Traditional Chinese Medicine, Jinan, China.

Scientific Reports
|September 2, 2022
PubMed

Insights

This study identifies novel autophagy-related genes as potential diagnostic biomarkers for dilated cardiomyopathy (DCM). Machine learning and bioinformatics pinpointed eight key molecular markers and ten gene-targeted drugs for DCM, improving diagnostic and therapeutic strategies.

Area of Science:

  • Genomics and Bioinformatics
  • Cardiovascular Research
  • Molecular Biology

Background:

  • Dilated cardiomyopathy (DCM) is characterized by impaired ventricular remodeling and function, often leading to heart failure and arrhythmias with a poor prognosis.
  • Identifying reliable diagnostic biomarkers for DCM is crucial for timely intervention and improved patient outcomes.
  • Autophagy-related genes (ARGs) play a significant role in cardiac health and disease, making them potential targets for DCM research.

Purpose of the Study:

  • To identify autophagy-related genes (ARGs) that can serve as diagnostic biomarkers for dilated cardiomyopathy (DCM).
  • To leverage machine learning and bioinformatics approaches for the discovery of novel DCM biomarkers.
  • To explore potential gene-targeted drugs and associated comorbidities for DCM management.

Main Methods:

  • Differential expression analysis of whole gene microarray data from the Gene Expression Omnibus (GEO) database (GSE4172 dataset).
  • Integration of autophagy gene libraries (HADb, HAMdb) to identify autophagy-related differentially expressed genes (AR-DEGs) in DCM.
  • Application of machine learning algorithms, R language for correlation analysis, Enrichr for pathway analysis, JASPAR for gene regulatory network construction, and DisGeNET for comorbidity analysis.

Main Results:

  • Identification of 23 autophagy-related differentially expressed genes (AR-DEGs) in DCM.
  • Screening of eight key molecular markers (PLEKHF1, HSPG2, HSF1, TRIM65, DICER1, VDAC1, BAD, TFEB) for DCM diagnosis using machine learning.
  • Establishment of a transcription factor gene regulatory network and identification of 10 gene-targeted drugs and associated comorbidities for DCM.

Conclusions:

  • The study successfully identified novel ARGs with potential as diagnostic biomarkers for DCM.
  • The identified molecular markers and gene-targeted drugs offer promising avenues for improving DCM diagnosis and treatment strategies.
  • Further validation is warranted to translate these findings into clinical practice for DCM patient care.