Time-Series Chlorophyll Fluorescence Imaging Reveals Dynamic Photosynthetic Fingerprints of sos Mutants to Drought Stress

• Published in: Sensors (Basel, Switzerland) Switzerland), 2019-06, Vol.19 (12), p.2649
• Main Authors: Sun, Dawei, Zhu, Yueming, Xu, Haixia, He, Yong, Cen, Haiyan
• Format: Article
• Language: English
• Subjects: Abiotic stress; Agricultural production; Arabidopsis thaliana; Automation; Biological activity; Chlorophyll; chlorophyll fluorescence imaging; Discriminant analysis; Drought; drought stress; Fingerprints; Fluorescence; Function analysis; Genes; Genomes; Genomics; Homeostasis; Kinases; Laboratories; Oxidative stress; Photosynthesis; Physiology; Proteins; Salt; salt overly sensitive (SOS) pathway; Signal transduction; Support vector machines
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s19122649

Resistance to drought stress is one of the most favorable traits in breeding programs yet drought stress is one of the most poorly addressed biological processes for both phenomics and genetics. In this study, we investigated the potential of using a time-series chlorophyll fluorescence (ChlF) analysis to dissect the ChlF fingerprints of salt overly sensitive (SOS) mutants under drought stress. Principle component analysis (PCA) was used to identify a shifting pattern of different genotypes including mutants and wild type (WT) Col-0. A time-series deep-learning algorithm, sparse auto encoders (SAEs) neural network, was applied to extract time-series ChlF features which were used in four classification models including linear discriminant analysis (LDA), k-nearest neighbor classifier (KNN), Gaussian naive Bayes (NB) and support vector machine (SVM). The results showed that the discrimination accuracy of mutants SOS1-1, SOS2-3, and wild type Col-0 reached 95% with LDA classification model. Sequential forward selection (SFS) algorithm was used to obtain ChlF fingerprints of the shifting pattern, which could address the response of mutants and Col-0 to drought stress over time. Parameters including , and , etc. were significantly different between mutants and WT. This research proved the potential of ChlF imaging for gene function analysis and the study of drought stress using ChlF in a time-series manner.