Manhattan Inputs and Outputs

Manhattan Inputs and Outputs#

To visualize results from the PheWAS, the PyPheWAS-package has a command called “make_manhattan” that allows users to generate a Manhattan plot of the results.

Manhattan Inputs:#

Input Arguments

  • –input-file, -i: Input file with the results from the PheWAS. This file should be tab-separated and have the columns ‘phecode_category’, ‘phecode_description’, and ‘converged’.

  • –output-file, -o: Filepath to output the Manhattan plot to. Default value is “test.png”.

  • –pval-col: Name of the column that has the p-values for the variable of interest. Default value is “pval”.

  • –beta-col: Name of the column that has betas for the variables of interest. Default value is “beta”.

  • –dpi: Quality of the image to output. Default value is 300.

Output Plot:#

Notable features of the output plot are described below:

  • Different phecode categories are illustrated by different colors on the x-axis

  • Bonferroni significance threshold is indicated by the red dashed line

  • Infinity lines are indicated by the dashed blue lines. In some PheWASes, the results can be so significant that they get rounded to zero. To account for this, the PyPheWAS-package finds the most significant non-zero result and multiplies that value by 1.02 to replace all zero values. The infinity lines are drawn at this non-zero value to indicate which values were replaced.

  • Positive and negative betas are indicated by the direction of the symbols on the plots