2016/07/29 by Jessica Hargreaves, Marina I. Knight, Hargreaves, Jessica K. +5 · 1 citation
Agricultural and Biological Sciences · Computer Science · #Applications (stat.AP) #FOS: Computer and information sciences #Plant and Biological Electrophysiology Studies #Time Series Analysis and Forecasting
paper · pdf · doi:10.48550/arxiv.1607.08827
openalex publication_date 2016/07/29 · openalex created_date 2022/10/07 · openalex updated_date 2026/07/28
How does soil pollution affect a plant's circadian clock? Are there any\ndifferences between how the clock reacts when exposed to different\nconcentrations of elements of the periodic table? If so, can we characterise\nthese differences?\n We approach these questions by analysing and modelling circadian plant data,\nwhere the levels of expression of a luciferase reporter gene were measured at\nregular intervals over a number of days after exposure to different\nconcentrations of lithium.\n A key aspect of circadian data analysis is to determine whether a time series\n(derived from experimental data) is `rhythmic' and, if so, to determine the\nunderlying period. However, our dataset displays nonstationary traits such as\nchanges in amplitude, gradual changes in period and phase-shifts.\n In this paper, we develop clustering methods using a wavelet transform.\nWavelets are chosen as they are ideally suited to identifying discriminant\nlocal time and scale features. Furthermore, we propose treating the observed\ntime series as realisations of locally stationary wavelet processes. This\nallows us to define and estimate the evolutionary wavelet spectrum. We can then\ncompare, in a quantitative way, using a functional principal components\nanalysis, the time-frequency patterns of the time series. Our approach uses a\nclustering algorithm to group the data according to their time-frequency\npatterns. We demonstrate the advantages of our methodology over alternative\napproaches and show that it successfully clusters our data.\n