eMalick G. Njie

Embarrassingly_FASTA: Enabling Recomputable, Population-Scale Pangenomics by Reducing Commercial Genome Processing Costs from $100 to less than $1

Embarrassingly_FASTA reduces genome processing costs from $100 to less than $1 by using GPU-accelerated preprocessing on ephemeral cloud infrastructure.

Seq2KING: An unsupervised internal transformer representation of global human heritages

Seq2KING uses transformer models to uncover genetic relatedness patterns in humans without predefined labels, revealing continent-based connectivity and population substructures.

dnaSORA - A Unified Diffusion Transformer for DNA point clouds

The paper introduces dnaSORA, a diffusion transformer for DNA point clouds, achieving state-of-the-art classification of C. elegans data at 0.3 Mb resolution using synthetic data.

Emergent populations derived with unsupervised learning of human whole genomes

This paper uses unsupervised machine learning to classify human genetic heritages and identifies previously unrecognized population strata in Japan.