Study IDs metabolites, genes as potential diagnostic markers in PAH
In-depth analyses, machine learning used to look for biomarkers of disease
Using in-depth analyses and machine learning, scientists have identified several molecules and genes involved in metabolism that may be diagnostic markers of pulmonary arterial hypertension (PAH).
Specifically, their research found five small molecules, or metabolites, and three metabolism-related genes that are tied to a PAH diagnosis. All were “identified using metabolomics, machine learning algorithms, and bioinformatics,” the researchers wrote. Metabolomics refers to the large-scale study of small molecules, commonly known as metabolites.
“Investigating the association between metabolic traits and PAH provides new insights into the [disease’s] underlying biological mechanisms, which could potentially improve the diagnosis and treatment of PAH,” the team wrote.
Their study, “Identification of metabolic biomarkers in idiopathic pulmonary arterial hypertension using targeted metabolomics and bioinformatics analysis,” was published in the journal Scientific Reports.
PAH is a rare type of pulmonary hypertension, which is marked by high blood pressure in the vessels that carry blood from the heart through the lungs. It’s usually idiopathic, meaning there isn’t an obvious underlying cause. An early diagnosis is key for optimal clinical outcomes in PAH, so there remains a continuing need to identify biomarkers that might be useful for detecting the disease.
Scientists find 5 molecules, 3 genes that may be diagnostic markers in PAH
In a search for PAH biomarkers, a trio of researchers in China conducted a series of analyses. The team used metabolomics, which in previous studies has shed light on potential biological pathways underlying PAH. Combining it with machine learning, a subset of artificial intelligence, has been successful in revealing metabolic markers of the disease.
Here, the scientists collected blood samples from 17 people with idiopathic PAH and 20 healthy people, who served as controls. The team then performed detailed analyses to measure levels of metabolites, which are the intermediate or end products of metabolism.
The metabolites then were analyzed using a series of machine learning algorithms. Machine learning involves feeding a dataset into a computer, alongside a set of algorithms that the computer can use to identify patterns in the data.
These analyses zeroed in on five metabolites that were present at significantly different levels in the PAH patients: adenosine monophosphate (AMP), homoserine, kynurenine, spermine, and tryptophan. The scientists noted that some of these metabolites are involved in the arginine pathway, a molecular pathway that’s important for making nitric oxide — a molecule that helps to regulate blood pressure by widening blood vessels.
To assess the utility of these five metabolites as potential diagnostic biomarkers, the researchers used a statistical calculation called the area under the receiver operating characteristic curve, or AUC. This measure assesses how well a particular parameter — in this case, biomarker levels — can distinguish between two groups, here, people with or without PAH. AUC values range from 0.5 to 1, with higher numbers indicating better accuracy. The AUC of these five metabolites was 0.952, suggesting a good ability to identify PAH.
Our findings revealed some key genes associated with metabolism in [pulmonary hypertension], … and may lead to the identification of useful metabolic biomarkers for the diagnosis of PAH.
In a separate series of analyses, the researchers looked at levels of different metabolism-related genes as potential PAH biomarkers. They found three genes that were expressed at significantly higher levels, meaning they were more active, in PAH. These were MAPK6, SLC7A11 and CDC42BPA. For these genes, AUC values in the identification of PAH all were higher than 0.8 in various datasets analyzed, again indicating generally good diagnostic capacity.
The researchers noted that these three genes are involved in some of the same biological pathways as the five metabolites. Among shared biological processes were immune response regulation and inflammation (kynurenine, tryptophan, and MAPK6), as well as cell growth (spermine and SLC7A11).
The scientists stressed that this study was based mainly on data from a small number of patients, so further work will be necessary to validate its findings. However, according to the team, these findings lay the foundation for further studies that ultimately may confirm new diagnostic biomarkers for PAH.
“Our findings revealed some key genes associated with metabolism in [pulmonary hypertension], and provided crucial information about complex metabolic reprogramming signals and may lead to the identification of useful metabolic biomarkers for the diagnosis of PAH,” the team wrote.
About the Author
