Identification of key genes in spontaneous cerebral hemorrhage and prevention of disease damage: LASSO and SVM regression

Prev Med. 2023 Sep:174:107633. doi: 10.1016/j.ypmed.2023.107633. Epub 2023 Jul 18.

Abstract

Prevention is more important than treatment, and the incidence of intracerebral hemorrhage can be effectively reduced by intervening on the risk factors of intracerebral hemorrhage. By studying the risk factors of spontaneous intracerebral hemorrhage, we can identify the risk factors to achieve the target of treatment and prevention. Through the use of the Least Absolute Shrinkage and Selection Operator (LASSO) and the Support Vector Machine (SVM), the two essential SICH-related genes, NUAK1 and ERO1L, were eliminated from consideration. A Venn analysis was performed, and based on the two important modules, it found that SICH was related with four critical genes: VCM1, CRNDE, COL6A2, and HSPB6. One gene (NUAK1) was dramatically downregulated in the illness group compared to the control group, whereas three essential genes (ERO1L, VCAM1, and COL6A2) were significantly upregulated in the disease group. In the end, the genes ERO1L, VCAM1, COL6A2, and NUAK1 were shown to be the most important ones for SICH. It is anticipated that these genes will become novel biomarkers as well as targets for the development of new pharmacotherapies for SICH.

Keywords: Key genes; Machine learning; Spontaneous intracerebral hemorrhage; WGCNA.

MeSH terms

  • Biomarkers
  • Cerebral Hemorrhage* / epidemiology
  • Cerebral Hemorrhage* / genetics
  • Humans
  • Protein Kinases
  • Repressor Proteins
  • Risk Factors
  • Support Vector Machine*

Substances

  • Biomarkers
  • NUAK1 protein, human
  • Protein Kinases
  • Repressor Proteins