Robust SNP-based prediction of rheumatoid arthritis through machine-learning-optimized polygenic risk score

BACKGROUND

The popular statistics-based Genome-wide association studies (GWAS) have provided deep insights into the field of complex disorder genetics. However, its clinical applicability to predict disease/trait outcomes remains unclear as statistical models are not designed to make predictions. This study employs statistics-free machine-learning (ML)-optimized polygenic risk score (PRS) to complement existing GWAS and bring the prediction of disease/trait outcomes closer to clinical application. Rheumatoid Arthritis (RA) was selected as a model disease to demonstrate the robustness of ML in disease prediction as RA is a prevalent chronic inflammatory joint disease with high mortality rates, affecting adults at the economic prime. Early identification of at-risk individuals may facilitate measures to mitigate the effects of the disease.

METHODS

This study employs a robust ML feature selection algorithm to identify single nucleotide polymorphisms (SNPs) that can predict RA from a set of training data comprising RA patients and population control samples. Thereafter, selected SNPs were evaluated for their predictive performances across 3 independent, unseen test datasets. The selected SNPs were subsequently used to generate PRS which was also evaluated for its predictive capacity as a sole feature.

RESULTS

Through robust ML feature selection, 9 SNPs were found to be the minimum number of features for excellent predictive performance (AUC > 0.9) in 3 independent, unseen test datasets. PRS based on these 9 SNPs was significantly associated with (P < 1 × 10^-16) and predictive (AUC > 0.9) of RA in the 3 unseen datasets. A RA ML-PRS calculator of these 9 SNPs was developed ( https://xistance.shinyapps.io/prs-ra/ ) to facilitate individualized clinical applicability. The majority of the predictive SNPs are protective, reside in non-coding regions, and are either predicted to be potentially functional SNPs (pfSNPs) or in high linkage disequilibrium (r2 > 0.8) with un-interrogated pfSNPs.

CONCLUSIONS

These findings highlight the promise of this ML strategy to identify useful genetic features that can robustly predict disease and amenable to translation for clinical application.

Unexplored Roles of Erythrocytes in Atherothrombotic Stroke

Stroke constitutes the second highest cause of morbidity and mortality worldwide while also impacting the world economy, triggering substantial financial burden in national health systems. High levels of blood glucose, homocysteine, and cholesterol are causative factors for atherothrombosis. These molecules induce erythrocyte dysfunction, which can culminate in atherosclerosis, thrombosis, thrombus stabilization, and post-stroke hypoxia. Glucose, toxic lipids, and homocysteine result in erythrocyte oxidative stress. This leads to phosphatidylserine exposure, promoting phagocytosis. Phagocytosis by endothelial cells, intraplaque macrophages, and vascular smooth muscle cells contribute to the expansion of the atherosclerotic plaque. In addition, oxidative stress-induced erythrocytes and endothelial cell arginase upregulation limit the pool for nitric oxide synthesis, leading to endothelial activation. Increased arginase activity may also lead to the formation of polyamines, which limit the deformability of red blood cells, hence facilitating erythrophagocytosis. Erythrocytes can also participate in the activation of platelets through the release of ADP and ATP and the activation of death receptors and pro-thrombin. Damaged erythrocytes can also associate with neutrophil extracellular traps and subsequently activate T lymphocytes. In addition, reduced levels of CD47 protein in the surface of red blood cells can also lead to erythrophagocytosis and a reduced association with fibrinogen. In the ischemic tissue, impaired erythrocyte 2,3 biphosphoglycerate, because of obesity or aging, can also favor hypoxic brain inflammation, while the release of damage molecules can lead to further erythrocyte dysfunction and death.

Machine learning-based prediction of the post-thrombotic syndrome: Model development and validation study

Background

Prevention is highly involved in reducing the incidence of post-thrombotic syndrome (PTS). We aimed to develop accurate models with machine learning (ML) algorithms to predict whether PTS would occur within 24 months.

Materials and methods

The clinical data used for model building were obtained from the Acute Venous Thrombosis: Thrombus Removal with Adjunctive Catheter-Directed Thrombolysis study and the external validation cohort was acquired from the Sun Yat-sen Memorial Hospital in China. The main outcome was defined as the occurrence of PTS events (Villalta score ≥5). Twenty-three clinical variables were included, and four ML algorithms were applied to build the models. For discrimination and calibration, F scores were used to evaluate the prediction ability of the models. The external validation cohort was divided into ten groups based on the risk estimate deciles to identify the hazard threshold.

Results

In total, 555 patients with deep vein thrombosis (DVT) were included to build models using ML algorithms, and the models were further validated in a Chinese cohort comprising 117 patients. When predicting PTS within 2 years after acute DVT, logistic regression based on gradient descent and L1 regularization got the highest area under the curve (AUC) of 0.83 (95% CI:0.76–0.89) in external validation. When considering model performance in both the derivation and external validation cohorts, the eXtreme gradient boosting and gradient boosting decision tree models had similar results and presented better stability and generalization. The external validation cohort was divided into low, intermediate, and high-risk groups with the prediction probability of 0.3 and 0.4 as critical points.

Conclusion

Machine learning models built for PTS had accurate prediction ability and stable generalization, which can further facilitate clinical decision-making, with potentially important implications for selecting patients who will benefit from endovascular surgery.

Possible mechanisms of cholesterol elevation aggravating COVID-19

Importance: Despite the availability of a vaccine against the severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2), humans will have to live with this virus and the after-effects of the coronavirus disease 2019 (COVID-19) infection for a long time. Cholesterol plays an important role in the infection and prognosis of SARS-CoV-2, and the study of its mechanism is of great significance not only for the treatment of COVID-19 but also for research on generic antiviral drugs. Observations: Cholesterol promotes the development of atherosclerosis by activating NLR family pyrin domain containing 3 (NLRP3), and the resulting inflammatory environment indirectly contributes to COVID-19 infection and subsequent deterioration. In in vitro studies, membrane cholesterol increased the number of viral entry sites on the host cell membrane and the number of angiotensin-converting enzyme 2 (ACE2) receptors in the membrane fusion site. Previous studies have shown that the fusion protein of the virus interacts with cholesterol, and the spike protein of SARS-CoV-2 also requires cholesterol to enter the host cells. Cholesterol in blood interacts with the spike protein to promote the entry of spike cells, wherein the scavenger receptor class B type 1 (SR-B1) plays an important role. Because of the cardiovascular protective effects of lipid-lowering therapy and the additional anti-inflammatory effects of lipid-lowering drugs, it is currently recommended to continue lipid-lowering therapy for patients with COVID-19, but the safety of extremely low LDL-C is questionable. Conclusions and Relevance: Cholesterol can indirectly increase the susceptibility of patients to SARS-CoV-2 and increase the risk of death from COVID-19, which are mediated by NLRP3 and atherosclerotic plaques, respectively. Cholesterol present in the host cell membrane, virus, and blood may also directly participate in the virus cell entry process, but the specific mechanism still needs further study. Patients with COVID-19 are recommended to continue lipid-lowering therapy.

LEAP-2: An Emerging Endogenous Ghrelin Receptor Antagonist in the Pathophysiology of Obesity

Liver-expressed antimicrobial peptide 2 (LEAP-2), originally described as an antimicrobial peptide, has recently been recognized as an endogenous blocker of growth hormone secretagogue receptor 1a (GHS-R1a). GHS-R1a, also known as ghrelin receptor, is a G protein-coupled receptor (GPCR) widely distributed on the hypothalamus and pituitary gland where it exerts its major functions of regulating appetite and growth hormone (GH) secretion. The activity of GHS-R1a is controlled by two counter-regulatory endogenous ligands: Ghrelin (activation) and LEAP-2 (inhibition). Ghrelin activates GHS-R1a on the neuropeptide Y/Agouti-related protein (NPY/AgRP) neurons at the arcuate nucleus (ARC) to promote appetite, and on the pituitary somatotrophs to stimulate GH release. On the flip side, LEAP-2, acts both as an endogenous competitive antagonist of ghrelin and an inverse agonist of constitutive GHS-R1a activity. Such a biological property of LEAP-2 vigorously blocks ghrelin's effects on food intake and hormonal secretion. In circulation, LEAP-2 displays an inverse pattern as to ghrelin; it increases with food intake and obesity (positive energy balance), whereas decreases upon fasting and weight loss (negative energy balance). Thus, the LEAP-2/ghrelin molar ratio fluctuates in response to energy status and modulation of this ratio conversely influences energy intake. Inhibiting ghrelin's activity has shown beneficial effects on obesity in preclinical experiments, which sheds light on LEAP-2's anti-obesity potential. In this review, we will analyze LEAP-2's effects from a metabolic point of view with a focus on metabolic hormones (e.g., ghrelin, GH, and insulin), and discuss LEAP-2's potential as a promising therapeutic target for obesity.

Association of four lipid components with mortality, myocardial infarction, and stroke in statin-naïve young adults: A nationwide cohort study

Aims

Dyslipidaemia is a modifiable cardiovascular risk factor with prognostic implications. Current strategies for lipid management in young adults are largely based on expert recommendations. We investigated the risks of death and cardiovascular disease in relation to each lipid component to establish evidence for primary prevention in young adults.

Methods

In this nationwide population-based cohort study, we analysed 5,688,055 statin-naïve subjects, aged 20–39 years, undergoing general health check-ups between 2009 and 2014. The endpoint was a composite of clinical events including death, myocardial infarction (MI), and stroke. We compared the incidence and risk of clinical events according to each lipid variable.

Results

During follow-up (median 7.1 years), clinical events occurred in 30,330 subjects (0.53%): 16,262 deaths (0.29%), 8578 MIs (0.15%), and 5967 strokes (0.10%). The risk of clinical events gradually increased with increasing total cholesterol (TC) and triglycerides and decreasing high-density lipoprotein cholesterol (HDL-C), largely driven by MI. Low-density lipoprotein cholesterol (LDL-C) had a J-shaped association with clinical events, showing the lowest risk for LDL-C of 84–101 mg/dL. Among lipid variables, triglycerides remained the sole independent predictor (adjusted hazard ratio, 1.20; p < 0.001) after adjusting for conventional risk factors.

Conclusions

For statin-naïve young adults, the risk of clinical events was proportional to lipid levels, positively with TC and triglycerides, negatively with HDL-C, and J-shaped with LDL-C. Triglycerides had an independent and the strongest association with the clinical events. Screening and intervention for abnormal lipid levels, particularly triglycerides, from an early age might be of clinical value.

PCSK9 inhibition and inflammation: A narrative review

Atherosclerotic cardiovascular disease (ASCVD) remains a leading cause of morbidity and mortality despite excellent pharmacological and revascularization approaches. Low-density lipoproteins (LDL) are undoubtedly the most significant biochemical variables associated with atheroma, however, compelling data identify inflammation as critical for the maintenance of the atherosclerotic process, underlying some of the most feared vascular complications. Although its causal role is questionable, high-sensitivity C-reactive protein (hs-CRP) represents a major biomarker of inflammation and associated risk in CVD. While statin-associated reduced risk may be related to the lowering of both LDL-C and hs-CRP, PCSK9 inhibitors leading to dramatic LDL-C reductions do no alter hs-CRP levels. On the other hand, hs-CRP levels identify groups of patients with a high risk of CV disease achieving better ASCVD prevention in response to PCSK9 inhibition. In the FOURIER study, even in patients with extremely low levels of LDL-C, there was a stepwise risk increment according to the values of hs-CRP: +9% (<1 mg/L), +10.8% (1-3 mg/L) and +13.1% (>3 mg/L). Likewise, in the SPIRE-1 and -2 studies, bococizumab patients with hs-CRP> 3 mg/L had a 60% greater risk of future CV events. Most of the patients enrolled in the PCSK9 trials were on maximally tolerated statin therapy at baseline, and an elevated hs-CRP may reflect residual inflammatory risk after standard LDL-C lowering therapy. Moreover, data on changes in inflammation markers in carriers of PCSK9 loss-of-function mutations are scanty and not conclusive, thus, evidence from the effects of anti-inflammatory molecules on PCSK9 levels might help unravel this hitherto complex tangle.