Rapid and accurate phenotyping of embryonic lethality via image analysis of C. elegans developmental stages from high-throughput image data Rapid and accurate phenotyping of embryonic lethality via image analysis of C. elegans developmental stages from h

WHITE A, KAO, CIPRIANI PG, MUNARRIZ E, PAABY A, GEIGER D, SONTAG E, GUNSALUS KC AND F. PIANO

Los Angeles

Congreso; 18th Internacional C. elegans Meeting; 2011

We present an automated image analysis system (DevStaR) for quantitative phenotyping of C. elegans embryonic lethality and sterility phenotypes. Our image analysis system counts each developmental stage in an image of a C. elegans population, allowing efficient high throughput calculation of C. elegans viability phenotypes. DevStaR is an object recognition machine comprising several hierarchical layers that build successively more sophisticated representations of the objects (developmental stages) to be classified. The algorithm segments the objects, decomposes the objects into parts, extracts features from these parts, and classifies them using an SVM (support Vector Machine) and global shape information. This enables correct classifications in the presence of complicated occlusions and deformations of the animals. Features of the classified objects are then used to obtain a count of each developmental stage. We are currently using this system to analyze phenotypic data from C. elegans high-throughput genetic screens, and have processed over one million images for lab users so far. Validation of DevStaR measurements will be shown by comparing DevStaR output to both manual counting of developmental stages and manualscores of quantitative phenotypes. DevStaR can provide an accurate measurement of quantitative phenotype and is comparable to manual scoring. DevStaR has been used to score a C. elegans genome wide RNAi screen with up to 30 repeats per clone tested at up to 5 temperatures per clone. The screen consists of over 600,000 images each scored by DevStaR, Analysis of these data illustrate the convenience of DevStaR scoring and the use of a quantitative phenotype. Our system overcomes a previous bottleneck in image analysis by achieving near real-time scoring of image data in a fully automated manner. Our system reduces the need for human evaluation of images and provides rapid quantitative output that is not feasible at high throughput by manual scoring.