PANorama: A Novel High Throughput Image-Processing Pipeline for Quantifying Panicle Architecture in Oryza sativa

Variation in inflorescence architecture within grasses (the Poaceae) is an evolutionary curiosity and one of the hallmarks of plant domestication. In Asian rice (Oryza sativa) the inflorescence is classified as a conically shaped compound raceme, also known as a panicle; spikelets and seeds form on long lateral branches, instead of directly on the main rachis axis. Panicle development in O. sativa occurs via a spatiotemporally coordinated transition between four types of meristems, which happens within a 7-day window known as panicle initiation (PI). Ultimately, the size and shape of a panicle has considerable consequences on physiology and yield performance, as spikelet number and position within a panicle determines seed carrying capacity and affects grain-filling quality. A number of candidate genes affecting panicle development and morphology have been identified using mutant analyses and are currently the target of modern breeding programs. However, due to the branched nature of an O. sativa panicle, it is still difficult to quantify and compare multiple aspects of inflorescence architecture. With the advent of modern genomics, it is becoming increasingly important to develop accurate phenotyping methodologies to leverage the growing body of high-resolution genotypic data sets. In order to quantitatively measure multiple aspects of panicle architecture, we have developed a phenotyping platform known as PANorama. PANorama is a high throughput pipeline that uses semi-automated computer processing and user-driven input to define and extract phenotypes from panicle images. We are currently analyzing panicle data generated with PANorama from several greenhouse trials.