Amber Emmett

Machine Learning to Predict Genetic Regulation in Blood Cells

About me

I am a bioinformatician undertaking my PhD at the University of Leeds, under supervision from Professor David Westhead. I received my undergraduate degree in Genetics from the University of Liverpool, where in my final year I researched the role of Ap4A in the DNA damage response in Fanconi Anaemia cells.

My project

I am currently working to develop machine learning models to predict regulatory elements involved in blood cell development.

Non-coding DNA comprises 99 % of the human genome. Whilst previously dismissed as ‘junk DNA’, we now know the non-coding genome is home to hundreds of thousands of regulatory sequences. Enhancers are cis-regulatory elements which upregulate gene expression by binding activator transcription factors at gene promoters. Unlike promoters, enhancers are located distal from their target gene. As a result, enhancers can be challenging to identify and characterise.

Machine learning can predict cis-regulatory elements from known genetic, epigenetic and structural features of these regions. This information comes from large, heterogeneous sets of–omics data, produced by next-generation sequencing methods like ChIP-seq, ATAC-seq, RNA-seq and Hi-C. My project applies machine learning techniques to these datasets, with the aim of predicting novel enhancers driving haematopoietic differentiation.


Not yet available.