caspo 1.0

Learning of Protein Signaling Logic Models powered by BioASP and CellNOptR

caspo combines BioASP and CellNOpt to provide an easy to use software for learning Protein Signaling Logic Models from a Prior Knowledge Network in .sif format and a phospho-proteomics dataset in MIDAS format.

Installation

You can install caspo by running:

$ pip install caspo

caspo will try to install the R package CellNOpt (R must be already installed). Note that you may need root access for this. Otherwise, you can install CellNOpt manually from the R console and use a virtualenv to install caspo.

Usage

Typical usage is:

$ caspo.py pkn.sif midas.csv

For more options you can ask for help as follows:

$ caspo.py --help
Usage: caspo.py [options] pkn.sif midas.csv

Options:
  -h, --help           show this help message and exit
  -t T, --tolerance=T  Suboptimal enumeration tolerance (Default to 0)
  -p P, --discrete=P   Discretization range exponent: 10^P (Default to 2)
  -q Q, --alpha=Q      Size penalty exponent 1/10^Q (Default to 5)
  -g, --gtts           Compute Global Truth Tables (Default to False). This
                       could take some time for many models.
  -o O, --outdir=O     Output directory path (Default to current directory)

Output

By default, the output of caspo will be 4 comma-separated-values files::

• models.csv: Matrix representation of logic models

• frequencies.csv: Frequencies of hyperedges occurrence

• exclusives.csv: Mutual exclusives hyperedges with their corresponding frequencies

• inclusives.csv: Mutual inclusives hyperedges with their corresponding frequencies

When using the -g option, caspo will also output::

• gtt_stats.csv: Basic cluster analysis.

• gtt-%i.csv: Explicit computation of each Global Truth Table