Spav

A tool for visualizing Splotch results.

Overview

A custom version of Spav is used on https://als-st.nygenome.org.

Spav currently supports five different interactive views for visualizing and exploring Splotch results.

Expression on arrays

This view visualizes expression estimates on all the arrays simultaneously.

Expression on array

This view visualizes expression estimates on a selected array. The user can filter arrays based on the level 1 information (see Splotch).

Expression in common coordinate

This view visualizes expression estimates in common coordinate system. The expression estimates are separated based on the level 1 information.

Expression coefficients per level 1 variable

This view visualizes expression coefficients across level 1 variables for each AAR.

Expression coefficients per anatomical annotation region (AAR)

This view visualizes expression coefficients across AARs for each level 1 variable.

Usage

Installation

Spav has been tested on Python 3.7.

Spav can be installed as follows

$ pip install git+https://git@github.com/tare/Spav.git

Data preparation

The script spav_prepare_data can be used for preparing the Splotch results to be used with Spav

$ spav_prepare_data --help
usage: spav_prepare_data [-h] -d DATA_DIRECTORY -o OUTPUT_DIRECTORY -s
                         SERVER_DIRECTORY [-c] [-v]

A script for preparing Splotch results for Span

optional arguments:
  -h, --help            show this help message and exit
  -d DATA_DIRECTORY, --data-directory DATA_DIRECTORY
                        data directory
  -o OUTPUT_DIRECTORY, --output-directory OUTPUT_DIRECTORY
                        output directory
  -s SERVER_DIRECTORY, --server-directory SERVER_DIRECTORY
                        server directory
  -c, --no-copy         create symbolic links instead of copying images
  -v, --version         show program's version number and exit

For instance, if your Splotch input and output directories are $DATA_DIRECTORY and $OUTPUT_DIRECTORY (see Splotch), respectively, then you can prepare the data for Spav by executing the following command

$ spav_prepare_data -d $DATA_DIRECTORY -o $OUTPUT_DIRECTORY -s $SPAV_DIRECTORY

This command will create the directories $SPAV_DIRECTORY/static and $SPAV_DIRECTORY/data. The directory $SPAV_DIRECTORY/static contains symbolic links pointing to the bright-field images and the $SPAV_DIRECTORY/data.hdf5 file contains the estimates.

Deployment

Standalone Bokeh server

For detailed description of how run a Bokeh server, please see https://bokeh.pydata.org/en/latest/docs/user_guide/server.html.

The script server/main.py implements our Bokeh application

$ python main.py --help
usage: main.py [-h] [--arrays] [--array] [--common-coordinate]
               [--aar-coefficients] [--level-coefficients] [-v]

Spav server

optional arguments:
  -h, --help            show this help message and exit
  --arrays              show the array view
  --array               show the arrays view
  --common-coordinate   show the common coordinate view
  --aar-coefficients    show the aar coefficient view
  --level-coefficients  show the level coefficient view
  -v, --version         show program's version number and exit

Additionally, the directory server contains theme.yaml and templates/index.html.

First, let us copy the files from the directory server to the directory we created using the spav_prepare_data script

$ cp -r server/* $SPAV_DIRECTORY/

Second, the Bokeh application with all the implemented views can started by executing the following command

$ bokeh serve $SPAV_DIRECTORY --show --args --arrays --array --common-coordinate --aar-coefficients --level-coefficients

Embedding the Bokeh server inside a Jupyter notebook

Please see Notebook.ipynb.

tare / Spav