Brett Allen McKinney

Gene Age Gap Estimate (GAGE) for major depressive disorder: a penalized biological age model using gene expression

The paper develops a gene expression-based model (GAGE) to estimate biological age gaps in MDD and finds significant associations with age and disease status.

Multivariate optimization of k for k-nearest-neighbor feature selection with dichotomous outcomes: complex associations, class imbalance, and application to RNA-Seq in Major Depressive Disorder

The paper develops and evaluates methods for optimizing k in k-nearest-neighbor feature selection for imbalanced data, showing improved performance on RNA-Seq data from Major Depressive Disorder.

Theoretical properties of nearest-neighbor distance distributions and novel metrics for high dimensional bioinformatics data

The paper derives analytical expressions for pairwise distance distributions under Lq metrics and introduces new metrics for genetic variants and rs-fMRI data to improve nearest-neighbor methods in bioinformatics.

Consensus Features Nested Cross-Validation

The paper introduces cnCV, a feature selection method combining nested cross-validation with differential privacy's feature stability, showing it is efficient and effective for classification with fewer false positives.

Nearest-neighbor Projected-Distance Regression (NPDR) for detecting network interactions with adjustments for multiple tests and confounding

The paper introduces NPDR, a feature selection method that detects network interactions in high-dimensional data while adjusting for multiple tests and confounding, showing improved performance over existing methods.

A humanized yeast phenomic model of deoxycytidine kinase to predict genetic buffering of nucleoside analog cytotoxicity

The paper uses a humanized yeast model to predict genetic buffering of nucleoside analog cytotoxicity by analyzing drug-gene interactions in cancer cell lines.

A nonlinear simulation framework supports adjusting for age when analyzing BrainAGE.

The paper proposes adjusting for chronological age in BrainAGE analysis to improve the detection of variable effects on brain aging.

Statistical Inference Relief (STIR) feature selection

The paper introduces STIR, a statistical inferential framework for Relief-based feature selection that accounts for variance in nearest neighbor distances to assess feature importance and significance in high-dimensional data.