Applying machine learning and artificial intelligence to decode complex patterns of DNA methylation


Epigenetic modifiactions, such as DNA methylation, play an important role in regulating gene expression. DNA methylation can also act as a “soft mutation”, creating phenotypic variation and allowing rapid adaption to environments. Decoding the complex patterns of methylation is a key barrier holding us back from understanding and harnessing this adaptability – particularly in response to climate change.

This project aims to take a data-driven approach to overcome the problem. We want to bring together large amounts of open data and develop machine learning algorithms that can identify methylation sites or spot patterns that are critical for regulating gene expression in multiple plant species.

The project is a collaboration between experts in epigenetics, genomics and bioinformatics at the Earlham institute and cutting-edge practitioners in machine learning and artifical intelligence at the John Innes Centre and IBM. This is an excellent PhD opportunity for a biologist to take a step into computational biology or for a data/computational scientist/mathematician to apply their knowledge to an important life science question.


Niederhuth CE, et al. (2016) Widespread natural variation of DNA methylation within angiosperms. Genome Biol 17(1):194.

L.-J. Gardiner, R. Joynson, J. Omony, R. Rusholme-Pilcher, L. Olohan, D. Lang, C. Bai, M. Hawkesford, D. Salt, M. Spannagl, K. F. X. Mayer, J. Kenny, M. Bevan, N. Hall, and A. Hall, “Hidden variation in polyploid wheat drives local adaptation.,” Genome Res,