vispr 0.2

Interactive HTML5 visualization for GeCKO screens.

VISPR is a visualization framework for CRISPR/Cas9 knockout screen experiments. To install VISPR, we recommend the Miniconda Python distribution (see below). Currently, VISPR is under heavy development and not yet recommended for public use.

Installing VISPR with the Miniconda Python distribution

Install Miniconda for Python 3 from here:

http://conda.pydata.org/miniconda.html

This will install a minimal Python 3, together with the conda package manager (if preferred, you can also use Python 2).

Then, open a terminal and execute

conda install -c johanneskoester vispr

to install VISPR with all dependencies. See below for running a test instance of VISPR.

Running VISPR

After successful installation, you can run VISPR by executing the command

vispr server path/to/config.yaml

See below for the config file format. If you only want to test VISPR, you can run a test instance with example data by executing

vispr test

Config file format (TODO update this)

The config file has to be in YAML format. It has to contain a section for each experiment. Each experiment contains two entries, one pointing to the target (or gene) result file provided by “MAGeCK test”, one pointing to a file with normalized read counts as provided by “MAGeCK count”:

myexperiment:

target_results: gene_summary.txt rna_counts: normalized_counts.txt

Installing VISPR with another Python distribution (for experts)

Make sure that you have numpy, scipy, scikit-learn and pandas installed. Else, their automatic compilation with the command below would take very long. Then, you can issue

pip install vispr

or

pip install vispr –user

if you want to install VISPR without admin rights. All remaining dependencies will be installed automatically.