mac389 / adunatos

gene-structure correlation in the human brain

adunatos

Quickstart

To create gene-structure matrix from Allen Brain Atlas

 cd src
 python access-aba.py

This module saves its output to as JSON and TXT. The module writes to the TXT file after each query. It dumps the JSON file after it has finished querying the Allen Brain Atlas's servers.

To isolate and compare regions of interest

 cd adunatos
 ./main.sh

This shell script calls four python files to:

1. Divide the gene-structure matrix into a control group that contains all brain areas and a regions of interest group (ROI), that contains all brain areas in which we are looking for enrichment. The file filter_words specifies which structures to include in the regions of interest group. The output is a Pandas DataFrame.

2. Annotate the DataFrame with the Uniprot IDs. The DataFrame initially only has Entrez gene names.

3. Annotate the DataFrame If your local machine does not recognize main.sh as executable, replace the last line with these two lines:

   chmod +x ./main.sh ./main.sh

To generate all data figures from Hawrylycz et al. (2012):

 ./make-all-figures.sh

To generate a specific figure:

 cd src
 python make-figure-x.py

(Built and tested in Python 2.7) #####Dependencies

• NumPy (built with 1.9.1)
• SciPy (built with 0.14.0)
• Matplotlib (1.4.0)
• Pandas (built with 0.14.1)
• mygene (built with 2.2.0)
• Awesome_print (0.1.3)