cy 0.1.7

Modelling CRISPR dropout data

Identify associations between genomic alterations (e.g. structural variation, copy-number variation) and CRISPR-Cas9 knockout response.

Description

Crispy uses Sklearn implementation of Gaussian Process Regression, fitting by default each chromosome of each sample independently.

Example

import pandas as pd
import crispy as cy

# Import data
data = cy.get_example_data()

# Association analysis
crispy = cy.CRISPRCorrection()\
    .fit_by(by=data['chr'], X=data[['cnv']], y=data['fc'])

# Export
crispy = pd.concat([v.to_dataframe() for k, v in crispy.items()])\
    .sort_values(['cnv', 'k_mean'], ascending=[False, True])

print(crispy)

Install

pip install cy

Enrichment and Regression module has Cython files, to compile run:

python crispy/enrichment/gsea_setup.py build_ext --inplace

python crispy/regression/linear_setup.py build_ext --inplace

Credits and License

Developed at the Wellcome Sanger Institue (2017-2018).

For citation please refer to: biorxiv pre-print - Tandem duplications lead to loss of fitness effects in CRISPR-Cas9 data