Fig. 2.

Phenotypic hit selection. a Number of direct segmentation phenotypes in each experiment identified as hits. The same small molecules may count multiple times, if phenotypes were observed in more than one treatment. b Dendrogram from hierarchical cluster analysis of phenotypic vectors. Colored boxes highlight clusters presented in c, e, g; high-resolution dendrogram in Supplementary Fig. 1. Correlation values above plot. c Direct segmentation phenotypes identified as “1st class phenotypes”, obtained from five different small molecules. d 10 µM SB225002 in wild type embryos caused specific mid-trunk defects of myotome boundaries (brackets, inset) with normal embryo morphology. e Direct segmentation phenotypes identified as “secondary tail appendage”. f 50 µM Wee1 Inhibitor in wild type embryos showing relatively normal morphology, but posterior segmentation defects (brackets, inset) with a small secondary tail appendage (arrow, dotted line). g, h Molecules were selected from cluster with segmentation defects and strong morphological defects. Temporally-controlled pulse in wild type embryos increased segmentation-specificity. g Standard long-term small molecule treatment with 50 µM estradiol caused strong developmental defects with rudimentary head and axis in wild type embryos (arrowheads). h Pulse experiments of 50 µM estradiol from 10 hpf for four hours showed recovered head and axis structures (arrowheads) and specific myotome boundary disruptions of up to 13 segments (bracket). Inset shows disrupted boundaries. Wild type embryos at 36 hpf, in situ hybridization for xirp2a. Scale bar: 200 µm