Introduction

This github repository contains the code underlying the CrispRVariantsLite Shiny-based web app, which accompanies the main CrispRVariants package. This web app is primarily available from the Robinson lab web server at University of Zurich from the following link: CrispRVariantsLite. Instructions are given in the start up page of the app.

Example data

Example ZIP files (and corresponding TXT files with description of genome location and/or guide+PAM sequence) for the different entry points can be found in the /examples/ directory. If retrieving these files from the github repo, be sure to click on 'Raw' or 'View Raw'.

Running this app locally

The CrispRVariantsLite web app can be run locally, after installation of the necessary R packages (see below) and well as the command line tools: samtools, bwa mem and the indices for the organism of interest. After all this is completed, clone a version of this repository, set up the symbolic links (see the makelinks file) in the data/genome and data/txdb directories of the cloned repository and then the app can be run from the R console as:

    #install.packages("shiny")  # only needed if 'shiny' package is not already installed
    shiny::runApp( "/dir/to/CrispRVariantsLite/local/clone")

Installing all the R packages

The following packages will be needed by the web app:

    source("https://bioconductor.org/biocLite.R")
    library(BiocInstaller)
    biocLite( c("CrispRVariants", "GenomicFeatures",
                "AnnotationDbi", "GenomicRanges",
                "IRanges", "Rsamtools", "Biostrings") )

    install.packages("ggplot2",quiet=TRUE)
    install.packages("shiny",quiet=TRUE)
    install.packages("shinydashboard",quiet=TRUE)
    install.packages("shinyBS",quiet=TRUE)
    install.packages("rhandsontable",quiet=TRUE)

Other important details

The web application makes calls to bwa mem and samtools. Therefore, for every organism of interest, a bwa index will need to be created from the genome sequence. These need to be copied (or preferably, symbolically linked) to the data/genome directory; similarly, TxDb objects need to be saved to the data/txdb directory of the app. The convention used is that all the files ending in .fa in the data/genome are presented to the user in the dropdown menu of the web application. Importantly, the TxDB objects in the data/txdb directory must be named with the same prefix but a .sqlite extension. That is, if you are working with the hg19 genome, the application is expecting hg19.fa (and hg19.fa.XXX for all the bwa indices) in the data/genome directory and hg19.sqlite in the data/txdb directory.

Pre-configured DOCKER container

You can also download and use a pre-configured docker container.

Docker Container

Please download the latest docker container version via

docker pull boutroslab/crisprvariantslite:latest

In order to get the genome reference, create a folder on your machine and paste all necessary genome files (as described in the makelinks file) into this folder.

e.g.

1. Download you genome reference build of choice from UCSC

2. Run BWA to create an index

   bwa index -a bwtsw GENOME.fa
   

3. Run SAMTOOLS to create an indexed FASTA

   samtools faidx GENOME.fa
   

4. Create the txdb sqlite file

   source("https://bioconductor.org/biocLite.R")
   library(GenomicFeatures)
   
   # e.g. for human genome hg38
   txdb <- makeTxDbFromUCSC("hg38", "knownGene")
   
   saveDb(txdb, file= "./genome/GENOME.sqlite")
   
   

Start CRISPRVariantsLite on your local machine using docker

Dont forget to map the genome folder via

-v PATHTOFOLDER:/srv/shiny-server/CRISPRVariantsLite/genome

docker run --rm -v PATHTOFOLDER:/srv/shiny-server/CRISPRVariantsLite/genome -p 3838:3838 boutroslab/crisprvariantslite:latest

How to Build the docker container

Clone or download the repository from https://github.com/jwinter6/CrispRVariantsLite/archive/master.zip

run docker build command within the docker subdirectory.