Using Machine Learning and Artificial Intelligence to Predict Diabetes Mellitus Among Women Population.

INTRODUCTION: Diabetes Mellitus (DM) is a chronic health condition (long-lasting) due to inadequate control of blood levels of glucose. This study presents a prediction of type 2 diabetes mellitus among women using various Machine Learning (ML) algorithms deployed to predict the diabetic condition. A University of California Irvine (UCI) diabetes mellitus dataset posted on Kaggle was used for analysis. METHODS: The dataset included eight risk factors for type 2 diabetes mellitus prediction, including age, systolic blood pressure, glucose, body mass index (BMI), insulin, skin thickness, diabetic pedigree function, and pregnancy. R language was used for the data visualization, while the algorithms considered for the study were logistic regression, Support Vector Machines (SVM), Decision Trees, and Extreme Gradient Boost (XGB). The performance analysis of these algorithms on various classification metrics was also presented, considering that the AUC-ROC score is the best for Extreme Gradient Boost (XGB) with 85%, followed by SVM and Decision Trees (DT). RESULTS: The Logistic Regression (LR) demonstrated low performance, but the decision trees and XGB showed promising performance against all the classification metrics. Moreover, SVM offers a lower support value, so it cannot be considered a good classifier. The model showed that the most significant predictors of type 2 diabetes mellitus were glucose levels and body mass index, whereas age, skin thickness, systolic blood pressure, insulin, pregnancy, and pedigree function were less significant. This type of real-time analysis has proven that the symptoms of type 2 diabetes mellitus in women fall entirely different compared to men, which highlights the importance of glucose levels and body mass index in women. CONCLUSION: The prediction of type 2 diabetes mellitus helps public health professionals to suggest proper food intake and adjust lifestyle activities with good fitness management in women to make glucose levels controlled. Therefore, the healthcare systems should give special attention to diabetic conditions in women. This work attempts to predict the occurrence of type 2 diabetes mellitus among women from their various behavioral and biological conditions.

Sociodemographic and Diabetes-related Risk Factors in San Diego County, California.

BACKGROUND: The research information would enable clinicians and public health professionals to formulate proper interventions for diabetic people according to age, gender, and race. OBJECTIVE: The aim of the study was to investigate the relationship between diabetes-related mortality, hospitalization and emergency department discharge, and sociodemographic characteristics, in addition to age-standardized mortality rate analysis. METHOD: A population-based cross-sectional descriptive study was carried out to determine the relationship between sociodemographic characteristics and diabetes-related risk factors of the San Diego County residents in 2018, including 49,283 individuals (27,366 males and 21,917 females). RESULTS: The outcomes were found to be statistically significant. Hospitalization and emergency department discharges among males and females were statistically significant. The statistical differences between gender and mortality were not significant. The mortality was not significant in the male group, while it was statistically significant in the female group. The noted agestandardized mortality rate of diabetes stood at 85.8 deaths per 100,000 standard population. CONCLUSION: This study found that mortality increases as people age, and 85% of deaths were found to be of people older than 65 years. The mortality was two times higher among white and Hispanic males than females. Findings from this study are important in understanding the sociodemographic characteristics at the county level, which can inform diabetes mortality prevention efforts.

Systemic computational investigation to identify potential inhibitors against cancer by targeting P21-activated kinase 4 and D(CGATCG).

Cancer accounts for more than 10 million deaths in the year 2020. Development of drugs that specifically target cancer signaling pathways and proteins attain significant importance in the recent past. The p21-activated kinase 4 enzyme, which plays diverse functions in cancer and is reported in elevated expression makes this enzyme an attractive anti-cancer drug target. Similarly, cancer cells' DNA could also serve as a good platform for anti-cancer drug development. Herein, a robust in silico framework is designed to virtually screen multiple drug libraries from diverse sources to identify potential binders of the mentioned cancer targets. The virtual screening process identified three compounds (BAS_01059603, ASN_10027856, and ASN_06916672) as best docked molecules with a binding energy score of ≤ -10 kcal/mol for p21-activated kinase 4 and ≤ -6 kcal/mol for D(CGATCG). In the docking analysis, the filtered compounds revealed stable binding to the same site to which controls bind in X-ray structures. The binding interactions of the compounds with receptors are dominated by van der Waals interactions. The average root mean square deviation (rmsd) value for p21-activated kinase 4 systems is noticed at ∼2 Å, while for D(CGATCG), the average rmsd is 2.7 Å. The MMGB/PBSA interpreted ASN_12674021 to show strong intermolecular binding energy compared to the other two systems and control in both receptors. Moreover, the entropy energy contribution is less than the mean binding energy. In short, the compounds are showing promising binding to the biomolecules and therefore must be evaluated for anti-cancer activity in experimental studies.Communicated by Ramaswamy H. Sarma.

Determination of Novel Anti-Cancer Agents by Targeting OGG1 Enzyme Using Integrated Bioinformatics Methods.

The 8-oxoguanine DNA glycosylase (OGG1) enzyme is a key DNA glycosylase mediating the excision of 7,8-dihydro-8-oxoguanine (8-oxoG) from DNA molecule to the start base excision repair pathway. The OGG1 glycosylase function depletion has been seen to obstruct pathological conditions such as inflammation, A3 T-cell lymphoblastic acute leukemia growth, and neurodegenerative diseases, thus warranting OGG1 as an attractive anti-cancer enzyme. Herein, we employed several drug libraries intending to screen non-toxic inhibitory molecules against the active pocket of the enzyme that achieved stable binding mode in dynamics. Two anti-cancer compounds ([O-]C1=C(CC2=CC=CC=C2)SC(=[N+]1CC(=O)NC3=NC=C(CC4=CC=CC=C4)S3)S and CCCN(CCC)[S]-(=O)(=O)C1=CC=C(C=C1)C(=O)NNC2=NC3=CC=C(Br)C=C3C(=N2)C4=CC=CC=C4) from Selleckchem.com were identified to occupy the active pocket of OGG1 and bind with greater affinity than Control TH5487. The binding affinity of Top-1 is -11.6 kcal/mol while that of Top-2 is -10.7 kcal/mol in contrast to TH5487 Control (-9 kcal/mol). During molecular dynamic simulations versus time, the said compounds are tightly held by the enzyme with no minor structural deviations reported except flexible loops in particular those present at the N and C-terminal. Both the compounds produced extensive hydrophobic interactions with the enzyme along with stable hydrogen bonding. The docking and molecular dynamics simulations predictions were further validated by molecular mechanics with generalized Born and surface area solvation (MM/GBSA) and Poisson Boltzmann surface area (MM/PBSA), and WaterSwap binding energies that validated strong binding of the compounds to the enzyme. The MM/GBSA binding free energy for Top-1 complex is -28.10 kcal/mol, Top-2 complex is -50.14 kcal/mol) and Control is -46.91 kcal/mol while MM/PBSA value for Top-1, Top-2 and Control is -23.38 kcal/mol, -35.29 kcal/mol and -38.20 kcal/mol, respectively. Computational pharmacokinetics support good druglike candidacy of the compounds with acceptable profile of pharmacokinetics and very little toxicity. All these findings support the notion that the compounds can be used in experiments to test their anti-cancer activities.