Audrey Kauffmann

MOBER integrates transcriptomic data from cancer cell lines, xenografts, and clinical tumors using deep learning to remove batch effects and improve clinical translatability.

The paper develops scAR, a machine learning model to remove ambient noise from single-cell omics data across multiple technologies, showing high denoising accuracy.

TRAWLING simplifies aberrant splicing detection by providing a splicing-aware transcriptome reference, improving RNAseq analysis across multiple datasets.

This paper introduces a fully unsupervised deep learning method to cluster cellular images by Mode-of-Action using only pixel intensity values, achieving high accuracy on a labeled dataset.

The paper introduces two methods to correct false positives in CRISPR screens caused by copy number variations, reducing false positives by 70-80% in amplified regions.

The paper identifies a subset of cancer cell lines that are misclassified as fibroblasts based on genetic and genomic analysis.