SPEER is a framework for predicting the tissue-specific effects of rare variants
Python Jupyter Notebook
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
simulated_data
src
src_output
tissue_groups
.gitignore
README.md

README.md

SPEER

SPEER (SPecific tissuE variant Effect predictoR) predicts tissue-specific regulatory effects of rare genetic variants using a hierarchial Bayesian model within a transfer learning framework. SPEER's advantages include:

  • integration of functional genomic annotations (from DNA sequence alone) with tissue-specific gene expression
  • separate predictions in each tissue while flexibly sharing information across tissues
  • computationally efficient algorithm that scales well to a large number of variants.

Installation

To download the code:

git clone https://github.com/farhand7/SPEER

SPEER is written in Python and requires the following packages: pandas, sklearn, numpy.

Usage

For a complete example of the SPEER pipeline using simulated data, see the ipython notebook

src/example.ipynb

For details on the SPEER algorithm, see

src/SPEER.py

To reproduce ROC curves using simulated data for all three settings described in the paper, see

src/simulate_data.py