Machine learning classification of plant genotypes grown under different light conditions through the integration of multi-scale time-series data

Abstract

In order to mitigate the effects of a changing climate, agriculture requires more effective evaluation, selection, and production of crop cultivars in order to accelerate genotype-to-phenotype connections and the selection of beneficial traits. Critically, plant growth and development are highly dependent on sunlight, with light energy providing plants with the energy required to photosynthesize as well as a means to directly intersect with the environment in order to develop. In plant analyses, machine learning and deep learning techniques have a proven ability to learn plant growth patterns, including detection of disease, plant stress, and growth using a variety of image data. To date, however, studies have not assessed machine learning and deep learning algorithms for their ability to differentiate a large cohort of genotypes grown under several growth conditions using time-series data automatically acquired across multiple scales (daily and developmentally). Here, we extensively evaluate a wide range of machine learning and deep learning algorithms for their ability to differentiate 17 well-characterized photoreceptor deficient genotypes differing in their light detection capabilities grown under several different light conditions. Using algorithm performance measurements of precision, recall, F1-Score, and accuracy, we find that Suport Vector Machine (SVM) maintains the greatest classification accuracy, while a combined ConvLSTM2D deep learning model produces the best genotype classification results across the different growth conditions. Our successful integration of time-series growth data across multiple scales, genotypes and growth conditions sets a new foundational baseline from which more complex plant science traits can be assessed for genotype-to-phenotype connections.

Keywords: Computer vision; Deep learning; Feature extraction; Machine learning; Phenotyping.

Graphical abstract

Fig. 1

: Lights treatments and plant phenotype. (A) Example depiction of the experimental light and camera setup. B) Example depiction of PlantCV data extraction. C) Twilights employed in the study. For the 0 min square bracket ramp, lights turned on to full Photosynthetic Photon flux Density (PPFD) or light intensity, while under the 30 min and 90 min conditions, the light intensity progressively increased to reach its maximum in 30 min and 90 min respectively. For all conditions, the plants received the exact amount of light of 4.32 DLI (mol/m²/d). (D) Spectral composition of the light for all treatments. (E) Plant list and associated genotypes. (F) Pictures of representative photoreceptor deficient plants grown under different twilight lengths. Plant position: 1 = WT, 2 = phyA, 3 = phyB, 4 = phyC, 5 = phyD, 6 = phyE, 7 = phot1, 8 = phot2, 9 = phot1/2, 10 = WT, 11 = cry1, 12 = cry2, 13 = cry1/2, 14 = cry3 ds-16, 15 = ds-16, 16 = fkf1, 17 = lkp2, and 18 = ztl.

Fig. 2

: The Residual Network’s architecture for time-series genotype classification of plants.

Fig. 3

: Encoder architecture for time-series genotype classification of plants.

Fig. 4

: ConvLSTM2D architecture for time-series genotype classification of plants.

Fig. 5

Performance of ConvLSTM2D model on the phototropin group (left) and the cryptochrome group (right) for different dawn and dusk periods (0, 30, and 90-minutes).

Fig. 6

Phototropin mutants details. (A) Pictures from the top of phototropin mutants phot1, phot2, and phot1/2 under different light treatments. (B) Pictures from the side of phot1, phot2 and phot1/2 under different light treatment.(C) Plant area in pixels of each phototropin mutant over 10 days calculated with PlantCV.