Pyhattan

Pyhattan is a package to created Manhattan plots from the results of GWA studies. It is designed to read text based files that include colums for chromosme, p_values, and, optionally, refSNP. Additional features include annotating SNP's exceeding a determined significance threshold and transforming p-values by -log10. Default parameters are set to read from a GEMMA output, but can be changed to work with any data meeting the required format.

Usage

FormatData(path, sep = '\t', chromosome = 'chr', p_value = 'p_wald')

• Description: Formats data from text file to be compatible with GenerateManhattan(). Requires column with chromosome id and p-value, optional- refSNP ID. Default parameters set to read from GEMMA output.

• path: path to data file.

• sep: data file deliminator.

• chromosome: [string] name of column that contains chromosome id.

• p_value: [string] name of column that contains p-values.

GenerateManhattan(pyhattan_object, export_path = None, significance = 6, colors = ['#E24E42', '#008F95'], refSNP = False)

• Description: Creates Manhattan plot from object created with FormatData() with significance line and significant SNP annotations

• pyhattan_object: pass through data.

• export_path: [string] optional path to save plot.

• significance: [float] signficance level, -log10(p-value).

• colors: [list] color pallettes.

• refSNP: [string] optional name of column that contains refSNP.

Example

Manhattan plot from GEMMA output

from Pyhattan import FormatData, GenerateManhattan

dat_gemma_output = FormatData('data/China_Pharm_Out.assoc.txt')
GenerateManhattan(dat_gemma_output,significance=2.5,refSNP='rs')

Manhattan plot from csv

dat_csv = FormatData('data/China_Pharm.csv',sep=',',chromosome='chromosome',p_value='p_value')
GenerateManhattan(dat_csv,significance=2.5, refSNP='refSNP')

Installation

Clone/Download the repo and enter the following in a command window:

        -------------------------------------------------------------------------------------
                python pip install ../Pyhattan/dist/Pyhattan-0.1-py2.py3-none-any.whl
        -------------------------------------------------------------------------------------

References

I base my code on this Stack Overflow answer by user Elad Joseph.