Epigenetic Signatures in Hypertension

Clear epigenetic signatures were found in hypertensive and pre-hypertensive patients using DNA methylation data and neural networks in a classification algorithm. It is shown how by selecting an appropriate subset of CpGs it is possible to achieve a mean accuracy classification of 86% for distinguis...

Full description

Saved in:
Bibliographic Details
Published in:Journal of personalized medicine 2023-05, Vol.13 (5), p.787
Main Authors: Alfonso Perez, Gerardo, Delgado Martinez, Victor
Format: Article
Language:English
Subjects:
Antihypertensives
Classification
Coronaviruses
COVID-19
DNA methylation
Epigenetic inheritance
Epigenetics
Hypertension
Kidney diseases
Machine learning
Methylation
Mortality
Neural networks
Patients
Precision medicine
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Clear epigenetic signatures were found in hypertensive and pre-hypertensive patients using DNA methylation data and neural networks in a classification algorithm. It is shown how by selecting an appropriate subset of CpGs it is possible to achieve a mean accuracy classification of 86% for distinguishing control and hypertensive (and pre-hypertensive) patients using only 2239 CpGs. Furthermore, it is also possible to obtain a statistically comparable model achieving an 83% mean accuracy using only 22 CpGs. Both of these approaches represent a substantial improvement over using the entire amount of available CpGs, which resulted in the neural network not generating accurate classifications. An optimization approach is followed to select the CpGs to be used as the base for a model distinguishing between hypertensive and pre-hypertensive individuals. It is shown that it is possible to find methylation signatures using machine learning techniques, which can be applied to distinguish between control (healthy) individuals, pre-hypertensive individuals and hypertensive individuals, illustrating an associated epigenetic impact. Identifying epigenetic signatures might lead to more targeted treatments for patients in the future.
ISSN:2075-4426
2075-4426
DOI:10.3390/jpm13050787