WorMachine: Machine Learning-Based Phenotypic Analysis Tool for Worms
While Caenorhabditis elegans nematodes are powerful model organisms, quantification of visible phenotypes is still often labor-intensive, biased, and error-prone. We developed \"WorMachine\", a three-step MATLAB-based image analysis software that allows automated identification of C. elegans worms, extraction of morphological features, and quantification of fluorescent signals. The program offers machine learning techniques which should aid in studying a large variety of research questions. We demonstrate the power of WorMachine using five separate assays: scoring binary and continuous sexual phenotypes, quantifying the effects of different RNAi treatments, and measuring intercellular protein aggregation. Thus, WorMachine is a \"quick and easy\", high-throughput, automated, and unbiased analysis tool for measuring phenotypes.