Incorporating prior knowledge into regularized regression.

MOTIVATION: Associated with genomic features like gene expression, methylation and genotypes, used in statistical modeling of health outcomes, there is a rich set of meta-features like functional annotations, pathway information and knowledge from previous studies, that can be used post hoc to facilitate the interpretation of a model. However, using this meta-feature information a priori rather than post hoc can yield improved prediction performance as well as enhanced model interpretation. RESULTS: We propose a new penalized regression approach that allows a priori integration of external meta-features. The method extends LASSO regression by incorporating individualized penalty parameters for each regression coefficient. The penalty parameters are, in turn, modeled as a log-linear function of the meta-features and are estimated from the data using an approximate empirical Bayes approach. Optimization of the marginal likelihood on which the empirical Bayes estimation is performed using a fast and stable majorization-minimization procedure. Through simulations, we show that the proposed regression with individualized penalties can outperform the standard LASSO in terms of both parameters estimation and prediction performance when the external data is informative. We further demonstrate our approach with applications to gene expression studies of bone density and breast cancer. AVAILABILITY AND IMPLEMENTATION: The methods have been implemented in the R package xtune freely available for download from https://cran.r-project.org/web/packages/xtune/index.html.

Preoperative hemoglobin A1c levels and increased risk of adverse limb events in diabetic patients undergoing infrainguinal lower extremity bypass surgery in the Vascular Quality Initiative.

BACKGROUND: The impact of preoperative glycemic control on the risk of adverse perioperative outcomes in diabetic patients undergoing lower extremity bypass (LEB) surgery is not well-understood. We determined whether higher preoperative hemoglobin A1c (HbA1c) levels are associated with an increased risk of major adverse limb events, major adverse cardiovascular events, and mortality in diabetic patients undergoing infrainguinal LEB. METHODS: A retrospective review of all infrainguinal LEB surgeries in the Vascular Quality Initiative registry from January 2012 to February 2017 was performed. Only surgeries performed on diabetic patients with complete demographic and clinical data, including HbA1c value at the time of LEB, were included for analysis (n = 7727). Entries were stratified according to the following HbA1c levels: 6 or less (n = 1087), greater than 6 to 7 or less (n = 2137), greater than 7 to 8 or less (n = 1657), and greater than 8 (n = 2846). Multivariate logistic regression was used to determine the association of preoperative HbA1c levels on the risk of in-hospital major adverse limb events (above ankle amputation, loss of primary graft patency), major adverse cardiovascular events (myocardial infarction, stroke, congestive heart failure, cardiac arrhythmia), and mortality. RESULTS: The number of surgeries complicated by adverse limb and cardiovascular events were 356 (4.6%) and 1314 (17.0%), respectively. There were 72 in-hospital deaths (0.9%). After adjustment for clinical and demographic variables, patients with high HbA1c values (≥8%) were at an increased risk of adverse limb events (odds ratio [OR], 1.37; 95% confidence interval [CI], 1.01-1.86) compared with those with a normal HbA1c (>6% to ≤7%). High HbA1c values were not associated with an increased risk of cardiovascular events (OR, 1.07; 95% CI, 0.81-1.43) or mortality (OR, 1.57; 95% CI, 0.83-3.03). Patients with low HbA1c values (≤6%) did not experience a significantly higher risk for any of the three outcomes. In a stratified analysis, the association of high HbA1c values with adverse limb events was only present in those presenting without critical limb ischemia (OR 1.82; 95% CI, 1.05-3.16). CONCLUSIONS: Poor preoperative glycemic control in diabetic individuals undergoing infrainguinal LEB, particularly in those without critical limb ischemia, is associated with an increased risk of in-hospital limb events. Further study should evaluate whether improved efforts to identify individuals with poorly controlled diabetes and subsequent interventions to better optimize glycemic control during the preoperative period improve limb outcomes after LEB.

Developmental trajectories for young children with 16p11.2 copy number variation.

Copy number variation at 16p11.2 is associated with diverse phenotypes but little is known about the early developmental trajectories and emergence of the phenotype. This longitudinal study followed 56 children with the 16p11.2 BP4-BP5 deletion or duplication between the ages of 6 months and 8 years with diagnostic characterization and dimensional assessment across cognitive, adaptive, and behavioral domains. Linear mixed modeling revealed distinct developmental trajectories with deletions showing VIQ gains but declines in motor and social abilities while duplications showed VIQ gains and steady development across other domains. Nonparametric analyses suggest distinct trajectories and early cognitive abilities for deletion carriers who are ultimately diagnosed with intellectual disability and developmental coordination disorder as well as distinct trajectories and early social communication and cognitive abilities for duplication carriers diagnosed with ASD and intellectual disability. Findings provide predictions for patient developmental trajectories, insight into mean functioning of individuals with 16p11.2 at early ages, and highlight the need for ongoing monitoring of social and motor functioning and behavioral symptomatology to improve treatment planning. © 2017 Wiley Periodicals, Inc.