MLb-LDLr

Potentials of artificial intelligence in familial hypercholesterolemia: Advances in screening, diagnosis, and risk stratification for early intervention and treatment

Familial hypercholesterolemia (FH) poses a global health challenge due to high incidence rates and underdiagnosis, leading to increased risks of early-onset atherosclerosis and cardiovascular diseases. Early detection and treatment of FH is critical in reducing the risk of cardiovascular events and improving the long-term outcomes and quality of life for affected individuals and their families. Traditional therapeutic approaches revolve around lipid-lowering interventions, yet challenges persist, particularly in accurate and timely diagnosis. The current diagnostic landscape heavily relies on genetic testing of specific LDL-C metabolism genes, often limited to specialized centers. This constraint has led to the adoption of alternative clinical scores for FH diagnosis. However, the rapid advancements in artificial intelligence (AI) and machine learning (ML) present promising solutions to these diagnostic challenges. This review explores the intricacies of FH, highlighting the challenges that are encountered in the diagnosis and management of the disorder. The revolutionary potential of ML, particularly in large-scale population screening, is highlighted. Applications of ML in FH screening, diagnosis, and risk stratification are discussed, showcasing its ability to outperform traditional criteria. However, challenges and ethical considerations, including algorithmic stability, data quality, privacy, and consent issues, are crucial areas that require attention. The review concludes by emphasizing the significant promise of AI and ML in FH management while underscoring the need for ethical and practical vigilance to ensure responsible and effective integration into healthcare practices.

OptiMo-LDLr: An Integrated In Silico Model with Enhanced Predictive Power for LDL Receptor Variants, Unraveling Hot Spot Pathogenic Residues

Familial hypercholesterolemia (FH) is an inherited metabolic disease affecting cholesterol metabolism, with 90% of cases caused by mutations in the LDL receptor gene (LDLR), primarily missense mutations. This study aims to integrate six commonly used predictive software to create a new model for predicting LDLR mutation pathogenicity and mapping hot spot residues. Six predictive-software are selected: Polyphen-2, SIFT, MutationTaster, REVEL, VARITY, and MLb-LDLr. Software accuracy is tested with the characterized variants annotated in ClinVar and, by bioinformatic and machine learning techniques all models are integrated into a more accurate one. The resulting optimized model presents a specificity of 96.71% and a sensitivity of 98.36%. Hot spot residues with high potential of pathogenicity appear across all domains except for the signal peptide and the O-linked domain. In addition, translating this information into 3D structure of the LDLr highlights potentially pathogenic clusters within the different domains, which may be related to specific biological function. The results of this work provide a powerful tool to classify LDLR pathogenic variants. Moreover, an open-access guide user interface (OptiMo-LDLr) is provided to the scientific community. This study shows that combination of several predictive software results in a more accurate prediction to help clinicians in FH diagnosis.

MLe-KCNQ2: An Artificial Intelligence Model for the Prognosis of Missense KCNQ2 Gene Variants

Despite the increasing availability of genomic data and enhanced data analysis procedures, predicting the severity of associated diseases remains elusive in the absence of clinical descriptors. To address this challenge, we have focused on the KV7.2 voltage-gated potassium channel gene (KCNQ2), known for its link to developmental delays and various epilepsies, including self-limited benign familial neonatal epilepsy and epileptic encephalopathy. Genome-wide tools often exhibit a tendency to overestimate deleterious mutations, frequently overlooking tolerated variants, and lack the capacity to discriminate variant severity. This study introduces a novel approach by evaluating multiple machine learning (ML) protocols and descriptors. The combination of genomic information with a novel Variant Frequency Index (VFI) builds a robust foundation for constructing reliable gene-specific ML models. The ensemble model, MLe-KCNQ2, formed through logistic regression, support vector machine, random forest and gradient boosting algorithms, achieves specificity and sensitivity values surpassing 0.95 (AUC-ROC > 0.98). The ensemble MLe-KCNQ2 model also categorizes pathogenic mutations as benign or severe, with an area under the receiver operating characteristic curve (AUC-ROC) above 0.67. This study not only presents a transferable methodology for accurately classifying KCNQ2 missense variants, but also provides valuable insights for clinical counseling and aids in the determination of variant severity. The research context emphasizes the necessity of precise variant classification, especially for genes like KCNQ2, contributing to the broader understanding of gene-specific challenges in the field of genomic research. The MLe-KCNQ2 model stands as a promising tool for enhancing clinical decision making and prognosis in the realm of KCNQ2-related pathologies.

Effect of mixed meal replacement of soybean meal on growth performance, nutrient apparent digestibility, and gut microbiota of finishing pigs

Introduction

This study was carried out to investigate the effects of mixed meal (rapeseed meal, cotton meal, and sunflower meal) replacement soybean meal on growth performance, nutrient apparent digestibility, serum inflammatory factors and immunoglobulins, serum biochemical parameters, intestinal permeability, short-chain fatty acid content, and gut microbiota of finishing pigs.

Methods

A total of 54 pigs with an average initial weight of 97.60 ± 0.30 kg were selected and randomly divided into 3 groups according to their initial weight, with 6 replicates in each group and 3 pigs in each replicate. The trial period was 26 days. The groups were as follows: control group (CON), fed corn-soybean meal type basal diet; Corn-soybean-mixed meal group (CSM), fed corn-soybean meal-mixed meal diet with a ratio of rapeseed meal, cotton meal, and sunflower meal of 1:1:1 to replace 9.06% soybean meal in the basal diet; Corn-mixed meal group (CMM), fed a corn-mixed meal diet with a ratio of Rapeseed meal, Cotton meal and Sunflower meal of 1:1:1 to replace soybean meal in the basal diet completely. The crude protein level of the three diets was maintained at 12.5%.

Results

Our findings revealed no significant impact of replacing soybean meal with the mixed meal (rapeseed meal, cotton meal, and sunflower meal) on the ADG (Average daily gain), ADFI (Average daily feed intake), and F/G (Feed gain ratio) (P > 0.05), or crude protein, crude fat, and gross energy (P > 0.05) in the diet of finishing pigs. Compared with the CON group, the serum interleukin 6 (IL-6) and interleukin 10 (IL-10) concentrations were significantly decreased in the CMM group (P < 0.05). However, there is no significant effect of the mixed meal (rapeseed meal, cotton meal, and sunflower meal) replacing soybean meal in the diet on the serum interleukin 1β (IL-1β), interleukin 8 (IL-8), tumor necrosis factor-alpha (TNF-α), immunoglobulin A (IgA), immunoglobulin G (IgG), and immunoglobulin M (IgM) concentrations (P > 0.05). Concordantly, there is no significant effect of mixed meal (rapeseed meal, cotton meal, and sunflower meal) replacing soybean meal in the diet on the serum antioxidant capacity, such as total antioxidant capacity (T-AOC), catalase (CAT), and malondialdehyde (MDA) levels of finishing pigs. Moreover, compared with the CON group, serum low-density lipoprotein (LDL-C) levels were significantly lower in the CSM group (P < 0.05) and their total bilirubin (TBIL) levels were significantly lower in the CMM group (P < 0.05). There is not a significant effect on serum D-lactate and diamine oxidase (DAO) concentrations (P > 0.05). The next section of the survey showed that the replacement of soybean meal with a mixed meal (rapeseed meal, cotton meal, and sunflower meal) in the diet did not significantly influence the acetic acid, propionic acid, butyric acid, valeric acid, isobutyric acid, and isovaleric acid in the colon contents (P > 0.05). Furthermore, compared with the CON group, the CMM group diet significantly increased the abundance of Actinobacteria at the phylum level (P < 0.05), U_Actinobacteria at the class level (P < 0.05), and U_Bacteria at the class level (P < 0.05). The result also showed that the CMM group significantly reduced the abundance of Oscillospirales at the order level (P < 0.05) and Streptococcaceae at the family level (P < 0.05) compared with the CON group. The Spearman correlation analysis depicted a statistically significant positive correlation identified at the class level between the relative abundance of U_Bacteria and the serum T. BILI concentrations (P < 0.05). Moreover, a significant negative correlation was detected at the order level between the relative abundance of Oscillospirales and the levels of acetic and propionic acids in the colonic contents (P < 0.05). Additionally, there was a significant positive correlation between the serum concentrations of IL-6 and IL-10 and the relative abundance of the family Streptococcaceae (P < 0.05).

Discussion

This study demonstrated that the mixed meal (rapeseed meal, cotton meal, and sunflower meal) as a substitute for soybean meal in the diet had no significant negative effects on the growth performance, nutrient apparent digestibility, serum immunoglobulins, serum antioxidant capacity, intestinal permeability, short-chain fatty acid content, and diversity of gut microbiota of finishing pigs. These results can help develop further mixed meals (rapeseed meal, cotton meal, and sunflower meal) as a functional alternative feed ingredient for soybean meals in pig diets.

Familial Hypercholesterolemia: From Clinical Suspicion to Novel Treatments

Familial hypercholesterolemia (FH) is the most common monogenic disorder in humans. It affects millions of people globally, increasing the risk of developing cardiovascular disease (CVD) at a younger age due to elevated levels of low-density lipoprotein cholesterol (LDL-C) from birth. While effective traditional and novel treatments are available, the most significant challenge with FH is the lack of timely diagnosis. As a result, many patients remain undertreated leading to an increased risk of CVD. To mitigate risk, initiating early and aggressive LDL-C-lowering therapies is recommended. Moreover, given its autosomal dominant inheritance patterns, it is also recommended to perform cascade lipid and/or genetic testing of all first-degree relatives. This review highlights the importance of early FH diagnosis and available treatment options. Greater awareness and improved screening efforts can help diagnose and treat more individuals, ultimately reducing the CVD risk associated with FH.

Applications of machine learning in familial hypercholesterolemia

Familial hypercholesterolemia (FH) is a common hereditary cholesterol metabolic disease that usually leads to an increase in the level of low-density lipoprotein cholesterol in plasma and an increase in the risk of cardiovascular disease. The lack of disease screening and diagnosis often results in FH patients being unable to receive early intervention and treatment, which may mean early occurrence of cardiovascular disease. Thus, more requirements for FH identification and management have been proposed. Recently, machine learning (ML) has made great progress in the field of medicine, including many innovative applications in cardiovascular medicine. In this review, we discussed how ML can be used for FH screening, diagnosis and risk assessment based on different data sources, such as electronic health records, plasma lipid profiles and corneal radian images. In the future, research aimed at developing ML models with better performance and accuracy will continue to overcome the limitations of ML, provide better prediction, diagnosis and management tools for FH, and ultimately achieve the goal of early diagnosis and treatment of FH.

Novel Tools for Comprehensive Functional Analysis of LDLR (Low-Density Lipoprotein Receptor) Variants

Familial hypercholesterolemia (FH) is an autosomal-dominant disorder caused mainly by substitutions in the low-density lipoprotein receptor (LDLR) gene, leading to an increased risk of premature cardiovascular diseases. Tremendous advances in sequencing techniques have resulted in the discovery of more than 3000 variants of the LDLR gene, but not all of them are clinically relevant. Therefore, functional studies of selected variants are needed for their proper classification. Here, a single-cell, kinetic, fluorescent LDL uptake assay was applied for the functional analysis of LDLR variants in a model of an LDLR-deficient human cell line. An LDLR-defective HEK293T cell line was established via a CRISPR/Cas9-mediated luciferase-puromycin knock-in. The expressing vector with the LDLR gene under the control of the regulated promoter and with a reporter gene has been designed to overproduce LDLR variants in the host cell. Moreover, an LDLR promoter-luciferase knock-in reporter system has been created in the human cell line to study transcriptional regulation of the LDLR gene, which can serve as a simple tool for screening and testing new HMG CoA reductase-inhibiting drugs for atherosclerosis therapy. The data presented here demonstrate that the obtained LDLR-deficient human cell line HEK293T-ldlrG1 and the dedicated pTetRedLDLRwt expression vector are valuable tools for studying LDL internalization and functional analysis of LDLR and its genetic variants. Using appropriate equipment, LDL uptake to a single cell can be measured in real time. Moreover, the luciferase gene knock-in downstream of the LDLR promotor allows the study of promoter regulation in response to diverse conditions or drugs. An analysis of four known LDLR variants previously classified as pathogenic and benign was performed to validate the LDLR-expressing system described herein with the dedicated LDLR-deficient human cell line, HEK293T-ldlrG1.

Continuous Bayesian variant interpretation accounts for incomplete penetrance among Mendelian cardiac channelopathies

PURPOSE

The congenital Long QT Syndrome (LQTS) and Brugada Syndrome (BrS) are Mendelian autosomal dominant diseases that frequently precipitate fatal cardiac arrhythmias. Incomplete penetrance is a barrier to clinical management of heterozygotes harboring variants in the major implicated disease genes KCNQ1, KCNH2, and SCN5A. We apply and evaluate a Bayesian penetrance estimation strategy that accounts for this phenomenon.

METHODS

We generated Bayesian penetrance models for KCNQ1-LQT1 and SCN5A-LQT3 using variant-specific features and clinical data from the literature, international arrhythmia genetic centers, and population controls. We analyzed the distribution of posterior penetrance estimates across 4 genotype-phenotype relationships and compared continuous estimates with ClinVar annotations. Posterior estimates were mapped onto protein structure.

RESULTS

Bayesian penetrance estimates of KCNQ1-LQT1 and SCN5A-LQT3 are empirically equivalent to 10 and 5 clinically phenotype heterozygotes, respectively. Posterior penetrance estimates were bimodal for KCNQ1-LQT1 and KCNH2-LQT2, with a higher fraction of missense variants with high penetrance among KCNQ1 variants. There was a wide distribution of variant penetrance estimates among identical ClinVar categories. Structural mapping revealed heterogeneity among "hot spot" regions and featured high penetrance estimates for KCNQ1 variants in contact with calmodulin and the S6 domain.

CONCLUSIONS

Bayesian penetrance estimates provide a continuous framework for variant interpretation.

Genetic Heterogeneity of Familial Hypercholesterolemia: Repercussions for Molecular Diagnosis

Genetics of Familial Hypercholesterolemia (FH) is ascribable to pathogenic variants in genes encoding proteins leading to an impaired LDL uptake by the LDL receptor (LDLR). Two forms of the disease are possible, heterozygous (HeFH) and homozygous (HoFH), caused by one or two pathogenic variants, respectively, in the three main genes that are responsible for the autosomal dominant disease: LDLR, APOB and PCSK9 genes. The HeFH is the most common genetic disease in humans, being the prevalence about 1:300. Variants in the LDLRAP1 gene causes FH with a recessive inheritance and a specific APOE variant was described as causative of FH, contributing to increase FH genetic heterogeneity. In addition, variants in genes causing other dyslipidemias showing phenotypes overlapping with FH may mimic FH in patients without causative variants (FH-phenocopies; ABCG5, ABCG8, CYP27A1 and LIPA genes) or act as phenotype modifiers in patients with a pathogenic variant in a causative gene. The presence of several common variants was also considered a genetic basis of FH and several polygenic risk scores (PRS) have been described. The presence of a variant in modifier genes or high PRS in HeFH further exacerbates the phenotype, partially justifying its variability among patients. This review aims to report the updates on the genetic and molecular bases of FH with their implication for molecular diagnosis.