Christopher Yau

The paper develops a cross-species deep learning framework to annotate and attribute ageing states in haematopoietic stem and immune cells, achieving high AUROC scores in human data.

The paper introduces a multi-omics integration framework that learns patient-specific embeddings from incomplete data using a variational autoencoder with a data-driven prior, achieving better performance on downstream tasks.

araCNA is a deep learning model that accurately predicts somatic copy number alterations in cancer genomes using simulated data and long-range sequence models.

GPerturb models single-cell perturbation data using Gaussian processes to estimate gene-level effects and provide uncertainty estimates, showing competitive performance on simulated and real datasets.

The paper uses single-cell sequencing to show that immune cells in the fallopian tube share clonotypes with ovarian cancer tumors, suggesting a role in early immune surveillance.

TransfoRNA uses an adaptive machine learning framework to improve small RNA annotation by iteratively expanding high-confidence labels and learning sequence-specific representations.

Rarity improves the detection of rare cell populations in single-cell imaging data through a Bayesian latent variable model.

The paper develops genetic scores for 17,227 molecular traits using multi-omic data from INTERVAL and validates them across ancestries, demonstrating their utility in pathway analysis and disease association discovery.

The paper develops genetically personalised organ-specific metabolic models to identify associations between metabolic fluxes and disease risk, finding 97 fluxes linked to coronary artery disease.

CIDER is a meta-clustering framework that improves scRNA-Seq data integration and evaluation by outperforming existing methods in simulated and real datasets without requiring prior cellular annotations.

The paper identifies four distinct endotypes in acute pancreatitis using multiomic data and clinical metrics, showing diverse molecular and systemic responses.

This study identifies patient-specific cancer genes in esophageal adenocarcinoma using machine learning, revealing recurrent process perturbations that suggest therapeutic targets.

The paper introduces Ouija, a method that infers pseudotime from marker genes to understand cell trajectory dynamics and detect metastable states.

This study performs a meta-analysis of PDAC gene expression data to identify immunologic transcriptional subtypes with prognostic significance.

The paper introduces a statistical framework to model pseudotime trajectories in heterogeneous genetic and environmental backgrounds, demonstrating its application to single-cell and population-level data.

This paper addresses biases in copy number-based deconvolution of tumour samples by introducing a simulation strategy to adjust for differential detection rates.