Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2014 Jun;26(6):2390-2403.
doi: 10.1105/tpc.114.124032. Epub 2014 Jun 10.

A Scalable Open-Source Pipeline for Large-Scale Root Phenotyping of Arabidopsis

Affiliations
Free PMC article

A Scalable Open-Source Pipeline for Large-Scale Root Phenotyping of Arabidopsis

Radka Slovak et al. Plant Cell. .
Free PMC article

Abstract

Large-scale phenotyping of multicellular organisms is one of the current challenges in biology. We present a comprehensive and scalable pipeline that allows for the efficient phenotyping of root growth traits on a large scale. This includes a high-resolution, low-cost acquisition setup as well as the automated image processing software BRAT. We assess the performance of this pipeline in Arabidopsis thaliana under multiple growth conditions and show its utility by performing genome-wide association studies on 16 root growth traits quantified by BRAT each day during a 5-d time-course experiment. The most significantly associated genome region for root growth rate is a locus encoding a calcium sensing receptor. We find that loss of function and overexpression of this gene can significantly alter root growth in a growth condition dependent manner and that the minor natural allele of the Calcium Sensor Receptor locus is highly significantly enriched in populations in coastal areas, demonstrating the power of our approach to identify regulators of root growth that might have adaptive relevance.

Figures

Figure 1.
Figure 1.
Multi-CCD Flatbed Scanner Cluster and the Image Acquisition Tool. (A) Eight Epson Perfection V600 photo scanners are used in parallel to provide high-throughput image acquisition. The scanners are operated by a single control unit. (B) Plate position is fixed with support frames. (C) To improve contrast, the scans are done in a dark room with the scanner lid open. (D) Screenshot of the software controlling parallel image acquisition.
Figure 2.
Figure 2.
Overview of BRAT Pipeline. (A) Whole plate image. Highlighted area corresponds to magnification shown in (B). Bar = 1 cm. (B) Segmented roots superimposed on the original image. Bar = 1 cm. (C) Flowchart of BRAT workflow. (D) Anatomy of a 3-d-old Arabidopsis seedling. (E) BRAT performance in supervised mode. (F) BRAT performance in unsupervised mode. FN, false negatives; TP, true positives; FP, false positives; N, total number of objects.
Figure 3.
Figure 3.
BRAT Is Robust toward Various Experimental Conditions. Images of whole plates and selected magnifications under different experimental conditions that lead to variations in image background color, image contrast, and the direction of root growth. Top row (from left to right): 1× MS medium, pH 5.7, 21°C; low temperature, 1× MS medium, pH 5.7, 10°C; high temperature, 1× MS medium, pH 5.7, 29°C; low pH, 1× MS medium, pH 4.6, 21°C; bottom row: sulfur-deficient medium, 1× MS medium − S, pH 5.7, 21°C; iron-deficient medium, 1× MS medium − Fe, pH 5.7, 21°C; phosphorus-deficient medium, 1× MS medium − P, pH 5.7, 21°C; auxin transport inhibitor (1-N-Naphtylphthalamic acid) treatment, 1× MS medium + 10 μM 1-N-naphtylphthalamic acid, pH 5.7, 21°C. White boxes in the whole plate images indicate an area that was magnified. Bars = 1 cm.
Figure 4.
Figure 4.
Genome-Wide Association Mapping for Root Traits. (A) to (D) Root linearity trait on day 3. (E) to (H) Root width at 20 to 40% interval of root on day 3. (I) to (M) Relative root growth rate trait (day 2 to day 3). Box indicates region depicted in (M). (A), (E), and (I) Examples for accessions with low trait values. (C), (G), and (K) Examples for accessions with high trait values. (B), (F), and (J) Histogram of mean trait values for accessions. (D), (H), and (L) Manhattan plots for GWAS. Line denotes 5% FDR threshold. Black box in (L) indicates location of most significant association of all GWAS performed in this study. (M) Genomic region surrounding the most significant GWA peak from (L). Top: gene models in genomic region. The x axis represents the position on chromosome 5. Bottom: −log10(P values) of association of the SNPs. Line denotes 5% FDR threshold. (N) and (O) Effects of loss of function and overexpression of CaS gene on root growth rate on day 7 to 8 (n > 26) on MS agar plates (pH 5.7) (N) and root length in hydroponic culture on day 5 (n = 17) in 2% MGRL nutrient solution (pH 5.0) (O). x axis, genotype; y axis, root growth rate (mm/day) or primary root length (mm), respectively; whiskers, ± 1.5 times the interquartile range; Student’s t test P value < 5*10−3 and < 5*10−4 indicated by two asterisks and three asterisks, respectively.
Figure 5.
Figure 5.
Minor Alleles of the CaS Gene Are Enriched in Coastal Populations of A. thaliana. (A) Geographic distribution of the top CaS SNP alleles in European accessions of the RegMap panel. CaS major alleles at the chromosome 5 position 7738620 (T) are depicted as blue symbols, whereas minor (A) alleles are depicted as red symbols. (B) Box plot of the distances from collection sites of accessions containing minor (red) or major (blue) alleles to the nearest coast. x axis, genotype; y axis, approximate distance to nearest coast (km); whiskers, ± 1.5 times the interquartile range.

Similar articles

See all similar articles

Cited by 39 articles

See all "Cited by" articles

LinkOut - more resources

Feedback