• Research

Exploring circadian rhythms in plants

Applying imaging techniques, transcriptomics and machine learning to explore circadian rhythms in a range of plant systems

Project summary.

Led by: Anthony Hall Group

Project Funders:

The circadian clock is a finely balanced time-keeping mechanism. This clock regulates a number of key cycles, from when we want to sleep to when we feel hungry. In plants, these rhythms are equally critical and regulate a huge range of processes, including flowering time, plant metabolism, and mineral uptake. 

We know that plants are healthier and have higher yields when their circadian clocks are in sync with their external surroundings. Understanding how the clock functions in key crops, such as bread wheat, has clear agricultural potential. 

Our group is interested in developing innovative methods for exploring clock regulation in different species and finding variation in circadian behaviour. For example, the impact of plants being the equivalent of a morning person (lark) or an evening person (owl). A core aim of this project is to translate techniques for studying circadian rhythms in Arabidopsis into agriculturally relevant species, with a focus on wheat. 


Delayed fluorescence

One way of profiling natural variation in circadian rhythms is by using delayed-fluorescence (DF) imaging. DF is light naturally re-emitted from photosynthetic tissues and so can be measured in any green plant without the need for genetic modification. So far, we have been able to use DF imaging to measure circadian rhythms in plants as varied as: Arabidopsis, Kalanchoe, liverworts, Brassicas, and wheat. We have adapted our DF method to optimise imaging of these plants. We recently used DF imaging to characterise circadian rhythms across 191 Arabidopsis ecotypes from Sweden and used Genome-Wide Association (GWA) to identify genes underpinning this variation. 

Machine learning

Another way we explore circadian regulation and variation is through computational approaches, such as machine learning (ML). In collaboration with researchers at IBM Research UK, we make use of publicly available transcriptional and genomic datasets to train, test, and validate ML models to answer a number of biological questions. So far, we have used ML approaches for:

  • classifying rhythmic data from transcriptomic datasets
  • strategic down-sampling of timepoints required to make rhythmic predictions 
  • making predictions about rhythmicity based on DNA-sequence alone
  • identifying biomarkers for predicting circadian time from a single transcriptomic time-point

Our current ML work is focused on defining a subset of genes to use as biomarkers to measure the internal circadian time within a plant. By comparing the relative expression of these marker genes, we can assay differences in clock behaviour as a result of natural variation, environmental stresses, or targeted genetic mutations. This work will give us the power to utilize publicly available transcriptomic datasets with just a single timepoint to look at differences in circadian phase. 

Circadian regulation in wheat

Another ongoing aspect of the project is focused on exploring how gene expression is coordinated in species with multiple copies of the same gene, as is the case in hexaploid wheat (AABBDD). By analysing circadian expression patterns in sets of three homoeologous genes (known as triads), we can identify cases where there are differential patterns in phase, period length, or rhythmicity in one of the homoeologs. Studying circadian regulation in polyploid crops is useful as it allows us to identify potential targets for manipulating clock function, which might have implications for yield and resilience. This work is in collaboration with the Antony Dodd group at the John Innes Centre.


Identifying marker genes to tell the circadian time using a single transcriptomic time point:



Interpreting machine learning models to investigate circadian regulation and facilitate exploration of clock function

Gardiner, L. J., Rusholme-Pilcher, R., Colmer, J., Rees, H., Crescente, J. M., Carrieri, A. P., Duncan, S., Pyzer-Knapp, E. O., Krishna, R., Hall, A., Designed, A. H., & Performed, S. D. (2021). 

Proceedings of the National Academy of Sciences of the United States of America, 118, 2021. 

Technology used.

These projects have benefited from sequencing and assistance from Genomic Pipelines, automation from the Earlham Biofoundry, and high-performance computing – all provided by the Earlham Institute.

We are also grateful for the support offered by Horticultural services and the Genome Resource Unit (SeedStor) at the John Innes Centre. 


John Innes Centre, Norwich

Antony Dodd Group

Prof. James Brown

IBM Research UK

Laura-Jayne Gardiner 

Impact statement.

We are constantly searching for new ways to improve wheat, a crop that provides more dietary calories than any other globally. By bringing our knowledge of circadian clocks in wheat in line with what we know for plants such as Arabidopsis, we can unlock further avenues for breeders and farmers to deploy environmentally robust and high yielding crops in the right places at the right times of year - maximising agricultural potential.