MaveVis: Visualization for MaveDB

A tool to visualize variant effect maps from MaveDB https://www.mavedb.org/ as genophenograms with added tracks for structure information. MaveVis is deployed as a webtool at http://vis.varianteffect.org.

Mavevis can be used in multiple different modes:

• A webtool, deployed as a docker container
• A RestFul webservice available from the same docker deployment
• An R-package
• A command line tool

Webtool

The webtool can be found at http://vis.varianteffect.org. But if you'd like to deploy it yourself you can get the built docker image at https://hub.docker.com/r/jweile/mavevis. You will of course need an installation of Docker. Then you can pull the image and deploy it as a container:

$ docker pull jweile/mavevis:latest

#Assuming we want to map the web interface to port 80:
$ docker run -t -p 80:80 --name mavevis jweile/mavevis:latest

If you insist on building the image from scratch, you can check out the git repo, and use the provided make file:

$ git clone https://github.com/VariantEffect/mavevis.git
$ cd docker
$ make build

RestFul service

A detailed manual of the webservice can be found here.

R-package

At the most basic level, MaveVis is available as an R-package, with the following dependencies:

Requires:

• R 3.1.2 or higher
• An installation of DSSP (tested with v3.0.0)
• An installation of FreeSASA (tested with v2.0.2)
• An installation of ClustalOmega (tested with v1.2.4)
• hgvsParseR
• rapimave
• yogitools
• httr
• gdata
• hash

Recommended:

• devtools
• testthat
• roxygen2
• cgir

Installation of the R-package:

Ubuntu/Debian/etc:

#Use apt to install meta-dependencies
$ sudo apt install r-base wget g++ make libcurl4-openssl-dev libssl-dev \
    libxml2 libxml2-dev libjson-c-dev dssp clustalo

#Download and build FreeSASA
$ wget https://github.com/mittinatten/freesasa/releases/download/2.0.2/freesasa-2.0.2.tar.gz
$ tar xzf freesasa-2.0.2.tar.gz
$ cd freesasa-2.0.2
$ ./configure --disable-xml
$ make
$ sudo make install

#Clean-up build directory
$ cd ..
$ rm -r freesasa*

#Install R-packages
$ R
> install.packages("devtools")
> library(devtools)
> install_github("jweile/yogitools")
> install_github("jweile/cgir")
> install_github("VariantEffect/hgvsParseR")
> install_github("VariantEffect/rapimave")
> install_github("VariantEffect/mavevis")
> q()

Using the R-package

A full manual for all functions included in the R-package can be found here.

CLI script

The main dashboard() function in the R-package can also be accessed as a CLI script, to avoid having to open an interactive session. You need to install the R-package as described above to use the script. You can then download the mavevis_launcher.R script from the docker directory and use it as follows:

Rscript mavevis_launcher.R scoresetID=<ssid> uniprot=<uniprot>
     pdb=<pdb-ids> mainChain=<chains> [WT=<seq> | seqOffset=<num> |
     synMed=<num> | stopMed=<num> | pngRes=<num> | outFormats={png|pdf|svg}]

The following parameters are mandatory: ssid is the URI of a score set in MaveDB, uniprot is the uniprot accession corresponding to the target of the scoreset, pdb is a comma-separated list of PDB accessions to use for the structure tracks, mainChain is a comma-separated list of PDB chain identifiers (e.g. A)

Depending on the scoreset, the following parameters may also be required: seqOffset the offset of the scoreset position indexes with respect to the uniprot sequence; synMed the expected score of a synonymous or wt-like variant; stopMed the expected score of a null-like or stop variant.

Finally, the following parameters are optional: WT is the WT sequence, which defaults to the one provided by MaveDB; outFormats is a comma-separated list of output file formats, with allowed values png, pdf, and svg; pngRes is the resolution (in DPI) to be used for any PNG output file.