Search for: All records

INTRODUCTION: It is unclear whether aggregated plasma protein risk scores (PPRS) could be useful to predict the risks of mild cognitive impairment (MCI) and Alzheimer’s disease (AD). METHODS: The Cox proportional hazard model with the LASSO penalty was used to build the PPRS for MCI and AD in 1,515 Framingham Heart Study Generation2 with 1,128 proteins measured in plasma at exam 5 [cognitive normal (CN)=1,258, MCI=129, AD=128]. RESULTS: MCI PPRS had a hazard ratio (HR) of 6.97[5.34,9.12], with a discriminating power (C-index=82.52%). AD PPRS had an HR of 5.74[4.67,7.05] (C-index=88.15%). Both PPRSs were also significantly associated with cognitive changes, brain-atrophy, and plasma AD biomarkers. Proteins in the MCI and AD PPRSs were enriched in several pathways related to leukocyte, chemotaxis, immunity, inflammation, and cellular migration. DISCUSSION: This study suggests that PPRS serve well to predict the risk of developing MCI and AD as well as cognitive changes and AD related pathogenesis in the brain. 

Description of endMT abnormalities and abnormal response to flow in TS iPSC-ECs 

Abstract Previous studies found many climate properties such as northern hemisphere (NH) surface temperature and precipitation respond non-monotonically when CO₂is increased from 1×to 8×CO₂relative to pre-industrial levels. Here, we explore the robustness of the non-monotonicity in the NH precipitation response in 11 coupled climate models. Eight models show a decrease in NH precipitation under repeated CO₂doubling, indicating that the non-monotonic response is a common but not universal result. Although common, the critical CO₂level where the NH precipitation decrease first occurs differs widely across models, ranging from 2×CO₂to 8×CO₂. These models also show a prominent weakening in the Atlantic meridional overturning circulation (AMOC) at the same critical CO₂level, with the AMOC weakening leading the precipitation decrease. The sensitivities of NH precipitation and the AMOC to CO₂doublings are positively correlated, especially when the AMOC weakens beyond 10 Sv. This suggests that the differences in models’ AMOC response can explain their contrasting NH precipitation responses, where models with a large AMOC weakening have decreased NH precipitation. Regionally, this decrease in NH precipitation is the most prominent over the North Atlantic, Europe and the tropical Pacific. Our results suggest that special care must be taken with the use of pattern scaling to inform regional climate decision-making. 

The expectation-maximization (EM) algorithm and its variants are widely used in statistics. In high-dimensional mixture linear regression, the model is assumed to be a finite mixture of linear regression and the number of predictors is much larger than the sample size. The standard EM algorithm, which attempts to find the maximum likelihood estimator, becomes infeasible for such model. We devise a group lasso penalized EM algorithm and study its statistical properties. Existing theoretical results of regularized EM algorithms often rely on dividing the sample into many independent batches and employing a fresh batch of sample in each iteration of the algorithm. Our algorithm and theoretical analysis do not require sample-splitting, and can be extended to multivariate response cases. The proposed methods also have encouraging performances in numerical studies.