hctsa_phenotypingWorm

A Matlab code package for running hctsa analysis on movement speed data of five strains of the nematode worm Caenorhabditis elegans: the CB4856 (Hawaiian wild isolate), and N2 (lab) strains, and the mutants dpy-20(e1282) (morphological mutant), unc-9(e101) (neural mutant), and unc-38(e264) (neural mutant).

The analysis results are published in this open-access article: 📗 Fulcher and Jones, Cell Systems (2017).

The data

The data were very kindly provided by Andre Brown and Bertalan Gyenes, Imperial College London, and can be downloaded, along with all of the hctsa calculations, either by running downloadComputedData, or from figshare. See wormbase for more information about the data, including gene knockouts. Note that the Hawaiian strain, CB4856, is labeled as H here.

hctsa feature extraction

In addition to the data file, HCTSA.mat, containing the results of hctsa feature extraction (on figshare), this repository also requires hctsa to be installed and paths to this package added (through the startup.m script in the hctsa package).

• wormAnalysis contains the main steps for processing and analysis.
• worm_TopFeatures allows more in-depth plotting for individual features.
• paperMainTextFigures reproduces the figures in the main text of the Cell Systems paper.
• doFilter is a function for filtering different sets of features and normalizing.

python analysis

Luiz Gustavo has worked up an analysis of the hctsa data using scikit-learn in python in this repository.