LDL estimation with equations and machine learning; variance of LDL and plasma atherogenic index

Cholesterol (CHOL), particularly low-density lipoprotein (LDL) CHOL, is responsible for many important diseases, including coronary artery disease, peripheral artery disease, and heart disease. In today’s modern society, this phenomenon has become a significant health issue for both adult...

Full description

Saved in:
Bibliographic Details
Main Authors: Petros Paplomatas, Marina Nikolidaki, Aristidis Vrahatis, Kostas Anagnostopoulos
Format: Article
Language:English
Published: Academia.edu Journals 2024-12-01
Series:Academia Molecular Biology and Genomics
Online Access:https://www.academia.edu/126364046/Estimation_of_Low_Density_Lipoprotein_LDL_values_using_equations_and_Machine_Learning_and_variance_calculation_of_LDL_and_Atherogenic_Index_of_Plasma_AIP_
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Cholesterol (CHOL), particularly low-density lipoprotein (LDL) CHOL, is responsible for many important diseases, including coronary artery disease, peripheral artery disease, and heart disease. In today’s modern society, this phenomenon has become a significant health issue for both adults and children, primarily due to dietary habits. LDL is a critical risk factor for atherosclerotic vascular disease or cardiovascular disease (CVD). However, the accurate determination of LDL is associated with high costs and technical difficulties such as ultracentrifugation. We present the LDLcalc package, which has two main functionalities: the first is the estimation of LDL from CHOL, high-density lipoprotein (HDL) CHOL, and triglycerides (TG). The second is the determination of the variance of LDL and atherogenic index of plasma (AIP), both of which are calculated clinical chemistry tests. LDL can be estimated using equations published in the literature or through machine learning (ML) methods. The package allows a comprehensive variety of approaches for estimating LDL values in patients, either through equations or through ten ML methods. When the direct measurement of LDL CHOL is not possible, equations or ML methods can be very good stand-ins. Additionally, the package provides the capability to calculate the variance of LDL and AIP using error propagation and bootstrap methods.
ISSN:3064-9765