Computer Finds Common Biomarkers for PAH, Metabolic Syndrome
Software tools also detected related metabolic pathways of both disorders
Genes associated with both pulmonary arterial hypertension (PAH) and metabolic syndrome have been identified using computer software tools, a study reports.
Metabolic syndrome, thought to promote PAH, is a cluster of conditions marked by high blood pressure, elevated blood sugar, excess body fat around the waist, and abnormal cholesterol or blood fat levels.
According to researchers, this is the first study to identify common biomarkers and related metabolic pathways of PAH and metabolic syndrome.
The computer study, “Identification of diagnostic biomarkers for idiopathic pulmonary hypertension with metabolic syndrome by bioinformatics and machine learning,” was published in the journal Nature Scientific Reports.
In PAH, the high blood pressure caused by the narrowing of the pulmonary arteries — the blood vessels that transport blood through the lungs — can be triggered by various diseases, including metabolic syndrome.
Although metabolic disorders are common among people with PAH, and growing evidence suggests a strong link between the two conditions, few studies have investigated them together.
One method to discover biological processes associated with a disease is to identify genes whose expression (activity) is different in diseased tissue compared with unaffected tissue. Such genes may participate in disease development and act as biomarkers to support diagnosis or monitor disease severity or treatment efficacy.
Gene Expression Omnibus (GEO) is a public database of functional genomic data submitted by the scientific community. Bioinformatics is a research field that uses computer software tools to analyze these large and complex datasets.
Research from China
Researchers at the Affiliated Hospital of Nantong University, in China, downloaded gene expression datasets from GEO related to PAH and metabolic syndrome and used bioinformatics to search for common diagnostic biomarkers.
Two datasets held lung tissue gene expression data collected from 50 PAH patient groups and 33 control groups, and one from 20 metabolic syndrome groups and 20 control groups, from blood samples.
The researchers first applied computer algorithms to screen for differentially expressed genes (DEGs) — differences in gene activity between patient and unaffected control samples — that were common between the two conditions.
In the combined PAH datasets, there were 159 DEGs, of which 88 were more active (upregulated) and 71 were less active (downregulated) than unaffected controls. In the metabolic syndrome dataset, 629 DEGs were upregulated, and 838 were downregulated. A comparison of the two results found 12 DEGs common to both conditions.
A second method, called weighted gene co-expression network analysis (WGCNA), was applied to the datasets to find clusters (groups) of highly correlated genes between the two conditions. From WGCNA, the team identified a further 280 PAH connections between metabolic syndrome genes. Among these, five matched with the 12 previously identified DEGs and were excluded from the analysis.
Biological processes related to the remaining 287 candidate genes indicated they were involved primarily in metabolism and immune responses and were closely related to the development of PAH and metabolic syndrome.
Because genes carry instructions to make proteins, which directly participate in disease biology, the researchers conducted a protein-protein interaction analysis to look for individual proteins that interact with several other proteins at so-called hubs. These hubs mainly played roles in immune responses, the team noted.
Last, a machine learning algorithm was applied to filter candidate genes of diagnostic value. This identified 11 genes with the highest value: EVI5L, RNASE2, PARP10, BSDC1, ACOT2, TMEM131, TNFRSF1B, SAC3D1, SLA2, P4HB, and PHF1.
Among these genes, EVI5L, with cell regulatory roles, RNASE2, associated with immune modulation, and PARP10, involved in fat metabolism, showed the highest predictive power.
Validating the findings
To validate these findings, the scientists examined the gene expression data from another dataset containing 17 lung tissue samples from eight PAH patient groups and nine control groups. Here, all candidate diagnostic genes were differentially expressed in lung tissue of PAH patients versus controls, with RNASE2 showing the most significant change.
Consistently, the team showed that PAH patients had elevated levels of immune cells associated with inflammatory immune responses than unaffected controls.
A “comprehensive analysis of the common biomarkers of these diseases can help with the early detection of hidden increased pulmonary vascular resistance in patients with [metabolic syndrome],” the researchers wrote, “with timely medical intervention enabling greater avoidance of serious consequences.”
“Interactions between said candidate diagnostic genes and dysregulated immune cells are still worth further studying,” the researchers noted.