Sam Kint

This study benchmarks DNA methylation deconvolution methods for tumoral fraction estimation in cfDNA, finding CelFiE most accurate and introducing DecoNFlow as a comprehensive pipeline.

This paper introduces MIMA, a deep learning framework for multi-omics integration and batch correction, which effectively removes batch effects and preserves biological information across modalities.

This study identifies two epigenetically-defined spatial niches in glioblastoma that influence cell state plasticity through local and long-range molecular signals.

The paper identifies club-like cells in the prostate tumor microenvironment that contribute to androgen deprivation therapy resistance through immune suppression.