A Bayesian Modelling Approach with Balancing Informative Prior for Analysing Imbalanced Data

Serum Zinc and Copper Concentrations and Dyslipidemia as Risk Factors of Cardiovascular Disease in Adults: Data Mining Techniques

This study aimed to examine the relationship between serum cholesterol levels and the ratio of zinc (Zn) and copper (Cu) in the blood serum and the incidence of cardiovascular disease (CVD). In Phase I of the study, 9704 individuals between the age of 35 and 65 years were recruited. Phase II of the cohort study comprised 7561 participants who completed the 10-year follow-up. The variables which were measured at the baseline of the study included gender, age, systolic blood pressure (SBP), diastolic blood pressure (DBP); biochemical parameters including serum Cu, Zn, copper-zinc ratio (Cu/Zn), zinc-copper ratio (Zn/Cu); fasted lipid profile consisting of triglycerides (TG), total cholesterol (TC), high-density lipoprotein (HDL), low-density lipoprotein (LDL) as well as fasting serum glucose, and triglycerides-glucose (TyG) index. Decision tree (DT) and logical regression (LR) models were applied to examine the relationship between the aforementioned factors and CVD. CVD was diagnosed in 837 individuals (378 males and 459 females) out of 7561 participants. According to the LR models, SBP, TC, HDL, age, Zn/Cu, and TyG index for males and SBP, age, TyG index, HDL, TC, Cu/Zn, and Cu for females had the highest correlation with CVD (p-value ≤ 0.033). Based on the DT algorithm, 88% of males with SPB < 129.66 mmHg, younger age (age < 53 years), TyG index < 9.53, 173 ≤ TC < 187 mg/dL, and HDL ≥ 32 mg/dL had the lowest risk of CVD. Also, 98% of females with SBP < 128 mmHg, TyG index < 9.68, age < 44, TC < 222 mg/dL, and HDL ≥ 63.7 mg/dL had the lowest risk of CVD. It can be concluded that the Zn/Cu for men and Cu/Zn for women, along with dyslipidemia and SBP, could significantly predict the risk of CVD in this cohort from northeastern Iran.

Predictive properties of novel anthropometric and biochemical indexes for prediction of cardiovascular risk

OBJECTIVE

Our aim was to examine the correlation between CVDs and various anthropometric and biochemical indices in the Iranian population.

METHODS

9704 healthy individuals without CVD aged 35-65 were enrolled in our study. The anthropometric indices including Body Adiposity Index (BAI), Abdominal Volume Index (AVI), Body Roundness Index (BRI), Waist to Hip Ratio (WHR), Weight-adjusted Waist Index (WWI), Conicity Index (C-Index), A Body Shape Index (ABSI), Waist to Height Ratio (WHtR), Body Surface Area (BSA), Body Mass Index (BMI), Lipid Accumulation Product (LAP) and Visceral Adiposity Index (VAI) were calculated. The biochemical indices including Cardiac Risk Ratio (CRR), Atherogenic Index of Plasma (AIP), Triglycerides-Glucose Index (TyG), Cardiac Risk Index (CRI), Atherogenic Coefficient (AC), and high-sensitivity C-Reactive Protein (hs-CRP) were investigated. The association of the above indices with CVD was analyzed using logistic regression (LR) and the decision tree (DT) models.

RESULTS

The LR showed age, hs-CRP, AIP, AVI, LAP, and TyG had significant associations with CVDs in men (p-value < 0.002). Also, age, hs-CRP, LAP, TyG, BRI, VAI, and CRR had significant associations with CVDs in women (p-value < 0.002). The DT showed 95% of men with age > = 48, AIP > = 0.94, TyG > = 9.71, and AVI > = 14.24 had CVDs. Also, 97% of women with age > = 54, TyG > = 8.33, and hs-CRP > = 36.69 had CVDs.

CONCLUSION

Age, TyG, AIP, AVI, hs-CRP and LAP were the best predictors of CVD in men. Moreover, age, TyG, hs-CRP and BAI were the best indicators of CVD in women.

Prediction of selective estrogen receptor beta agonist using open data and machine learning approach

Background

Estrogen receptors (ERs) are nuclear transcription factors that are involved in the regulation of many complex physiological processes in humans. ERs have been validated as important drug targets for the treatment of various diseases, including breast cancer, ovarian cancer, osteoporosis, and cardiovascular disease. ERs have two subtypes, ER-α and ER-β. Emerging data suggest that the development of subtype-selective ligands that specifically target ER-β could be a more optimal approach to elicit beneficial estrogen-like activities and reduce side effects.

Methods

Herein, we focused on ER-β and developed its in silico quantitative structure-activity relationship models using machine learning (ML) methods.

Results

The chemical structures and ER-β bioactivity data were extracted from public chemogenomics databases. Four types of popular fingerprint generation methods including MACCS fingerprint, PubChem fingerprint, 2D atom pairs, and Chemistry Development Kit extended fingerprint were used as descriptors. Four ML methods including Naïve Bayesian classifier, k-nearest neighbor, random forest, and support vector machine were used to train the models. The range of classification accuracies was 77.10% to 88.34%, and the range of area under the ROC (receiver operating characteristic) curve values was 0.8151 to 0.9475, evaluated by the 5-fold cross-validation. Comparison analysis suggests that both the random forest and the support vector machine are superior for the classification of selective ER-β agonists. Chemistry Development Kit extended fingerprints and MACCS fingerprint performed better in structural representation between active and inactive agonists.

Conclusion

These results demonstrate that combining the fingerprint and ML approaches leads to robust ER-β agonist prediction models, which are potentially applicable to the identification of selective ER-β agonists.