Additional file 4: Figure S2. of WorMachine: machine learning-based phenotypic analysis tool for worms
Adam Hakim
Yael Mor
Itai Toker
Amir Levine
Moran Neuhof
Yishai Markovitz
Oded Rechavi
10.6084/m9.figshare.5793084.v1
https://springernature.figshare.com/articles/presentation/Additional_file_4_Figure_S2_of_WorMachine_machine_learning-based_phenotypic_analysis_tool_for_worms/5793084
Features used to establish worm masculinity. Violin plots show the morphological features (A) and fluorescent features (B) used to determine the masculinity of him-5; [tph-1p::GFP] worms. A total of 545 pre-labeled worms of each sex were used for analysis (****p < 10â4, ***p < 10â3, *p < 0.05, two-tailed t test after Îą = 0.01 trimming to exclude extreme outliers with false discovery rate (FDR) correction for multiple comparisons). Only features that were significantly distinct and had plausible theoretical justification to differentiate between sexes were used for sex phenotype prediction. As can be seen in this figure, males and hermaphrodites differ in some features but not in every feature examined. (PPTX 597 kb)
2018-01-16 05:00:00
Caenorhabditis elegans
Machine learning
Deep learning
High-throughput image analysis
Feature extraction
Image processing
Phenotype analysis