Guillaume Marçais

The paper investigates the impact of using canonical k-mers in sketching methods for genomic sequences and proposes solutions to address resulting issues.

This paper introduces masked minimizers, a unifying framework for sequence sketching methods, and proposes a new metric to evaluate and optimize them for better genomic analysis.

The paper introduces polar sets for sequence-specific minimizer construction, showing their effectiveness on the human reference genome.

HARVESTMAN is a framework for hierarchical feature learning and selection from whole genome sequencing data, showing improved performance over binary SNP representations in breast cancer classification.

The paper introduces a randomized algorithm called Miniception to design minimizers with optimal density guarantees for practical bioinformatics applications.

The paper introduces a method to create practical universal k-mer sets for minimizer schemes that can handle larger k-mer lengths, reducing false positives in sequence matching.

The paper introduces an LSH method for edit distance that considers both k-mer content and order, improving alignment detection efficiency.

The paper analyzes minimizers schemes and shows that asymptotically optimal schemes with lower density than previously thought exist, leading to more efficient bioinformatics tools.

The paper improves minimizers and winnowing schemes by analyzing k-mer ordering and introducing universal hitting sets to enhance performance.