---
title: "vcfR workflow"
author: "Brian J. Knaus"
date: "`r Sys.Date()`"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{vcfR workflow}
  %\VignetteEngine{knitr::rmarkdown}
---



## Workflow


Work begins with a variant call format file (VCF).
A reference file (FASTA) and an annotation file are also suggested.
These later two files are not necessary, but I feel they contribute substantially.
The below flowchart outlines the major steps involved with vcfR use.



![flowchart image](vcfR_flowchart.png)



The green rectangles contain functions that the user will want to become familiar with.
There are effectively three phases in the workflow: reading in data, processing the objects in memory and plotting the data graphically.
Reading in the data frequently presents a bottleneck.
Unfortunately, this is typically due to the time it takes to read from a drive and therefore may not be anything I can improve on.
Once the data are in memory we can manipulate it.
VcfR uses Rcpp to implement functions written in C++ to try to improve the performance of these functions.
Lastly, we visualize the objects.
Visualization relies on R's base graphics package.
This is also something that I am not likely to be able to improve on if it becomes a bottleneck.
By dividing the analysis into these three phases I've divided that workflow into things I may be able to improve upon through my code writing, and things which I have little control over.