Michael David Morgan

Scarecrow improves data recovery from single-cell combinatorial barcoding by accounting for positional errors in barcode sequences, enhancing downstream analyses.

The paper identifies transcriptional noise as a driver of cancer treatment adaptation through AP-1 and KDM5-mediated chromatin rewiring in neuroblastoma and hepatoblastoma.

This study identifies two postnatal thymus epithelial progenitor cell subpopulations defined by Ly6d expression, showing differential impact of ageing on their maintenance.

This study identifies three modes of plasticity in neuroblastoma that shape tumour evolution under selective pressures and therapy resistance.

Milo2.0 enables population genetic analysis of cell state abundance using a count-based mixed model, identifying over 500 csQTLs and linking them to immune-mediated disease.

The paper investigates human primordial germ cell specification using in vitro models and in vivo datasets, revealing the role of TFAP2A and TFAP2C in PGC fate and amnion development.

Milo tests differential abundance in single-cell data using k-NN graphs to detect subtle cell state changes without relying on discrete clusters.

The paper maps the genetic architecture of hair colour in the UK population, identifying multiple loci and epistatic interactions contributing to red and blonde hair.

This paper introduces a batch correction method for scRNA-seq data that matches mutual nearest neighbours, showing improved performance over existing methods on simulated and real data.