PD Genetic Landscape Plot

• Author(s) - Frank Grenn
• Date Started - August 2020
• Quick Description: - R shiny application that takes in a specifically formatted csv file containing Parkinson's disease (or any other disease you have data for) summary statistics and plots variants by their effect sizes and frequencies.

File Overview

• ui.R - code for user interface layout.
• server.R - code for server logic.
• global.R - library imports.
• datafields.R - contains code to initialize some of the user interface. Also contains variables used to modify what the inputs display.
• risk_variant_data.ipynb - notebook containing code used to organize the main input file (risk_variant_data.csv) for this tool.
• www/labeler.js - code used to space the annotations on the plot.
• www/plot.js - code used to generate the plot.
• www/setup.js - contains variables that control display and the data the plot uses.
• www/risk_variant_data.csv - file containing all data to be displayed in the plot. Generated in risk_variant_data.ipynb.
• www/theme.css - css to format and style the display.

Making Changes

Some things to consider if modifying this tool yourself.

1) All data comes from www/risk_variant_data.csv

• this is read into the tool in server.R
• each row in this file contains a significant PD variant.
• ideally this file should at least contain columns for:
  • a unique identifier (rsid, chrbp, etc)
  • a frequency (MAF, effect allele frequency, etc)
  • an effect size (beta, odds ratio, etc)
• if you are changing this file, you will likely need to update datafields.R and www/setup.js (see steps below).

2) Input options can be changed in datafields.R

• the options for all inputs come from lists and dataframes initialized in this file.
• They are used to map names displayed by the input to actual field names used in the input .csv (ex: displayed "nearest gene" will use the field "NEAR_GENE" in the input .csv)

3) Plot display and additional data options can be changed in www/setup.js

• contains variables used for display, like the size of the plot and what to include in the tooltip.
• also contains objects used to map names to data fields.