Enabling Phenotypic Image Analysis Using Shared Cyberinfrastructure

Understanding how an organism's genotype and environment influence its growth, development and physiology requires an array of genetic and phenotyping tools. While many genetic resources exist including high-throughput sequencing methods, t-DNA insertion lines, gene micro-arrays, and structured genomic populations for statistical genomic studies, there are relatively few high-throughput  methods for monitoring and modeling dynamic and complex phenotypes. Presented here is an example of image processing methods applied to analyzing root gravitropism. High-spatiotemporal (5 μm/pixel and 2 min/frame) imaging of root gravitropism can be automatically analyzed via machine vision technologies and phenotypic features extracted including growth rate, tip angle, and curvature. These data-rich phenotypes can be combined with tensor algebra to produce data models which can succinctly describe a complex process with a small set of biologically relevant numbers. Making the phenotypic image analysis tools available as a shared resource can enable the plant community to more quickly make detailed phenotypic measurements. iPlant's cyber-infrastructure is a flexible platform able to deploy and house these computational methods as a shared community resource.  Currently, a growing community is leveraging these tools for phenotypic assay's and educational purposes.