Search

Frailty, a prevalent clinical syndrome in aging adults, is characterized by poor health outcomes, represented via a standardized frailty-phenotype (FP), and Frailty Index (FI). While the relevance of the syndrome is gaining awareness, much remains unclear about its underlying biology. Further elucidation of the genetic determinants and possible underlying mechanisms may help improve patients' outcomes allowing healthy aging. Genotype, clinical and demographic data of subjects (aged 60–73 years) from UK Biobank were utilized. FP was defined on Fried's criteria. FI was calculated using electronic-health-records. Genome-wide-association-studies (GWAS) were conducted and polygenic-risk-scores (PRS) were calculated for both FP and FI. Functional analysis provided interpretations of underlying biology. Finally, machine-learning (ML) models were trained using clinical, demographic and PRS towards identifying frail from non-frail individuals. Thirty-one loci were significantly associated with FI accounting for 12% heritability. Seventeen of those were known associations for body-mass-index, coronary diseases, cholesterol-levels, and longevity, while the rest were novel. Significant genes CDKN2B and APOE, previously implicated in aging, were reported to be enriched in lipoprotein-particle-remodeling. Linkage-disequilibrium-regression identified specific regulation in limbic-system, associated with long-term memory and cognitive-function. XGboost was established as the best performing ML model with area-under-curve as 85%, sensitivity and specificity as 0.75 and 0.8, respectively. This study provides novel insights into increased vulnerability and risk stratification of frailty syndrome via a multi-modal approach. The findings suggest frailty as a highly polygenic-trait, enriched in cholesterol-remodeling and metabolism and to be genetically associated with cognitive abilities. ML models utilizing FP and FI + PRS were established that identified frailty-syndrome patients with high accuracy. [ABSTRACT FROM AUTHOR]

Endometriosis is a condition characterized by implants of endometrial tissues into extrauterine sites, mostly within the pelvic peritoneum. The prevalence of endometriosis is under-diagnosed and is estimated to account for 5–10% of all women of reproductive age. The goal of this study was to develop a model for endometriosis based on the UK-biobank (UKB) and re-assess the contribution of known risk factors to endometriosis. We partitioned the data into those diagnosed with endometriosis (5924; ICD-10: N80) and a control group (142,723). We included over 1000 variables from the UKB covering personal information about female health, lifestyle, self-reported data, genetic variants, and medical history prior to endometriosis diagnosis. We applied machine learning algorithms to train an endometriosis prediction model. The optimal prediction was achieved with the gradient boosting algorithms of CatBoost for the data-combined model with an area under the ROC curve (ROC-AUC) of 0.81. The same results were obtained for women from a mixed ethnicity population of the UKB (7112; ICD-10: N80). We discovered that, prior to being diagnosed with endometriosis, affected women had significantly more ICD-10 diagnoses than the average unaffected woman. We used SHAP, an explainable AI tool, to estimate the marginal impact of a feature, given all other features. The informative features ranked by SHAP values included irritable bowel syndrome (IBS) and the length of the menstrual cycle. We conclude that the rich population-based retrospective data from the UKB are valuable for developing unified machine learning endometriosis models despite the limitations of missing data, noisy medical input, and participant age. The informative features of the model may improve clinical utility for endometriosis diagnosis. [ABSTRACT FROM AUTHOR]