Pharmacogenetics

INTRODUCTION

The PharmProfiler aims to provide a convenient way to extract pharmacogenetic data from your patient of choice. It includes a method to create a database based on the PharmGKB website, and methods to use this database to create a report for your individual.

PRE-REQUISITES

• You will require a PharmGKB account.
• Python 2.7 with the following packages:
  • BioPython >= 0.6.2
  • tqdm >= 4.19.4
  • requests_oauthlib >= 0.8.0
  • Flask
  • Clustalo
  • FastTree

BUILDING THE DATABASE

The following command can be used to get general input help:

python pharmprofiler.py -h

To create all necessary tables in the proper order, execute the following command:

python pharmprofiler.py -b all

This will prompt your PharmGKB e-mail address and password multiple times. You can also (re)create individual tables replacing '-t all' with various forms of -t YOURTABLE'. (options include: pairs, genes, variants, haplotypes, guidelines, annotations and more.) If you do this route, keep in mind that the following order is required (all valid options for the '-b' command).

pairs > drugs > drugmatches > drugvars > genes > haplotypes > hapvars > etcvars > indels > annotations > guidelines > guideoptions

This means that the table before the > is necessary to build the table after the >.

PRE-PROCESSING

In order to process patient data, a GVCF file is required. This file can be generated through the GATK pipeline. To execute this query, you will require a BED file. To create this file, please first complete database creation and then execute the following command.

python pharmprofiler.py -b bed

It will create a file that GATK can use to determine the gene locations to search in. Please execute the GATK VariantCaller with the options --emitRefConfidence 'BP_RESOLUTION' and -L BED_FILE_LOCATION to create the GVCF file needed for further steps.

IMPORTING GVCF

Once you have your GVCF file and your PharmProfiler database has been created, you can proceed with importing your individual's data. If you want to do everything in one row, including generating a report, execute the following:

python pharmprofiler.py -p GVCF_LOCATION -a all

You can also execute steps separately, from importing to haplotyping to finding matching annotations.

python pharmprofiler.py -p GVCF_LOCATION -a YOUR_ACTION

If you do this route, keep in mind that the following order is required (all valid options for the '-a' command).

import > haplotype > score > genotype > guidelines > annotate > report

This means that the action before the > is necessary to perform the action after the >.

REPORT generation

Using the command:

python pharmprofiler.py -p GVCF_LOCATION -a report

After executing all other patient options as listed above will net you a Latex file with a report in it. You will require TexMaker or another latex renderer to convert this to PDF. Other formats will be added in the future. ( ;) )

---

Copyright notice

Copyright 2016 Joanna Wolthuis & Joep de Ligt