Michaël G. B. Blum

DiagAI, a machine-learning system, improves variant interpretation in genetic testing by prioritizing causal variants, outperforming existing tools with higher diagnostic yield and shorter variant shortlists.

This study evaluates deep learning-based representation methods for survival and gene essentiality prediction on bulk RNA-seq data, finding that non-DL methods perform comparably on survival tasks while DL methods excel in gene essentiality prediction.

This paper develops a symptom interaction model to address heterogeneous phenotype reporting in genetic disease, improving gene prioritization by 42%.

The paper analyzes genetic structure in Brittany and the Loire basin to understand Western Europe's demographic history, finding increased population differentiation linked to steppe and Neolithic ancestry.

The paper presents best practices and efficient implementations for PCA of genetic data to address issues like LD structure capture, shrinkage bias, and population structure confounding.

This study evaluates DNA methylation deconvolution methods for cell-type proportion inference, identifies critical steps for accuracy, and develops a benchmark pipeline.

The paper improves polygenic score prediction by optimizing multiple hyperparameters in Clumping and Thresholding and introduces Stacked Clumping and Thresholding (SCT) for better performance.

This paper introduces an efficient penalized logistic regression method for genetic risk prediction, showing improved performance over existing methods like C+T and random forests.

Loter is a software package for local ancestry inference across species, showing robustness to admixture time and outperforming existing tools in simulations.

This paper evaluates redundancy analysis (RDA) for identifying loci under selection by comparing it to PCA and LFMM in simulations and applying it to Populus trichocarpa SNP data.

The paper introduces two R packages, bigstatsr and bigsnpr, for efficiently managing and analyzing large-scale genomic data within a single framework.

The paper evaluates how user expertise affects genome scan outcomes, finding that less experienced users can achieve similar results to advanced ones with training, but struggle with intermediate difficulty challenges.