Skip to content

carringtonlab/tcv-image-analysis

Repository files navigation

PlantCV-based Image Analysis of Arabidopsis Infection with TCV

Image Analysis with PlantCV

Parallelization with HTCondor

The script parallel-plantcv-arabidopsis-tcv.py was used to create an HTCondor job cluster submission file. The script walks through an input directory of images and creates a job for each image using the plantcv-arabidopsis-tcv-image.py script as the executable.

usage: parallel-plantcv-arabidopsis-tcv.py [-h] --dir DIR --pdfs PDFS --outdir
                                           OUTDIR --jobfile JOBFILE
                                           [--debug DEBUG]

Create an HTCondor job to process Arabidopsis images infected with TCV.

optional arguments:
  -h, --help         show this help message and exit
  --dir DIR          Directory containing images. (default: None)
  --pdfs PDFS        Naive Bayes PDF file. (default: None)
  --outdir OUTDIR    Output directory for images. (default: None)
  --jobfile JOBFILE  Output HTCondor job file. (default: None)
  --debug DEBUG      Activate debug mode. Values can be None, 'print', or
                     'plot' (default: None)

The script plantcv-arabidopsis-tcv-image.py is a PlantCV (http://plantcv.danforthcenter.org) analysis script. For each input image it outputs an image with plant pixels classified as "healthy" or "unhealthy" and an image with these classes transparently overlaid onto the original image. It also outputs several text files per image that can be concatenated to produce output tables for hue histograms, general hue statistics, and 1-, 2-, and 3-component Gaussian Mixture Models.

usage: plantcv-arabidopsis-tcv-image.py [-h] --image IMAGE --pdfs PDFS
                                        --outfile OUTFILE --outdir OUTDIR
                                        [--debug DEBUG]

Process Arabidopsis images infected with TCV.

optional arguments:
  -h, --help         show this help message and exit
  --image IMAGE      An image file. (default: None)
  --pdfs PDFS        Naive Bayes PDF file. (default: None)
  --outfile OUTFILE  Output text filename. (default: None)
  --outdir OUTDIR    Output directory for images. (default: None)
  --debug DEBUG      Activate debug mode. Values can be None, 'print', or
                     'plot' (default: None)

After the hue histogram data is concatenated into a single table, the R script hue_analysis.R was used to generate ridgeline plots for the pilot (Col-0 and dcl234) and complete (Col-0, dcl234, and ago1 through 10 single mutants).

Releases

No releases published

Packages

No packages published