Skip to content

Administration Guide - Quick Start

Quick-Start

From running instance to first filter query, the following steps need to be taken in-order:

  • Create a cooperation
  • Create a project
  • Create necessary diseases
  • Create a sample sheet
  • Upload the sample sheet
  • Annotate and upload your vcf files
  • Filter Autosomal (dominant|recessive) or de-novo

Once you started the application and have confirmed that it is in working order, you can start to administer your system. This includes setting up users and uploading a first dataset to filter for disease causing variants.

All of the following steps are described in greater detail in the different parts of this section. For the impatient, to be able to filter your first case, these are the minimal steps you need to do.

Evaluation

As there are many moving parts within EVAdb, especially around the annotation side of the application it is good practice to first upload a known solved case to test variant filtration and annotation.

Access

To follow this guide you need SSH access to the host running EVAdb and admin access to both applications (user and admin).

Create a Cooperation and Project

Prior to uploading samples, we need to setup the support structure. Goto New cooperation and enter your cooperation details in the form. Typically, this will be your name and details. Afterwards create a project and assign your new cooperation to the project.

We assume your cooperation is named Mustermann and the project is called EXOME_CASES.

Create diseases

For the sample upload and filter processes to work, we need to create some diseases. These tables are not pre-populated. Assuming you start with a single cohort of trio-samples it is advisable to set the parents to a different disease group than the index cases. Create at least two diseases, one "real" disease that can be attributed to one of the disease groups and one for the control samples that you assign the disease group Controls.

A simple, yet unspecific example could be two diseases:

  1. Developmental disorder, DEV (group: Developmental disorder) and
  2. Controls, CTRLS (group: Controls)

Setup a sample sheet

The sample sheet is the most convenient way to define multiple sample objects at once. While it is possible to create samples manually or via the LIMS system, we recommend using sample sheets as source of truth. Every row in the sample sheet defines exactly one sample and all its relevant attributes. Most of the columns are mandatory. Upload your sample sheet by using the Import external samples form.

Sample ID,Foreign ID,Pedigree,Comment,Sex,Affected,Organism,Tissue,Disease,Library Type,Read Type,Exome Assay
76436,76436,B20-0498,TranslateNamse,female,1,human,peripheral blood,Cases,exomic,paired-end,SureSelect60Mbv6
76443,76443,B20-0498,TranslateNamse,male,0,human,peripheral blood,Controls,exomic,paired-end,SureSelect60Mbv6
76442,76442,B20-0498,TranslateNamse,female,0,human,peripheral blood,Controls,exomic,paired-end,SureSelect60Mbv6

The three ID columns Sample ID, Foreign ID and Pedigree are used to identify samples. Additionally, since Pedigree is used as family id, when the same Pedigree is set for multiple samples, we will be able to define the family structure through the web user interface. To be able to use de-novo filter and the like it is necessary to specify the family structure by modifying the generated sample entries.

Disease

Make sure to set Disease column to the values you used in the previous steps.

Annotate and Upload your variants

Docker

The following part is specific to the docker version of EVAdb. When using a bare-metal or other installation please adjust the commands.

After defining samples and the family structure for your cohort, the next step is uploading the variants. The minimum prerequisite other than the previous step is that all data is local to the server that hosts the mariadb container.

Database

Make sure the database is running and able to handle connections using the MYSQL_USER username prior to starting the annotation container.

To make your data available inside the container, set DATA_DIR in your .env file to the directory containing all vcf files that you wish to upload. If you wish to annotate and insert a specific .vcf file, use the following procedure. We assume that the vcf file is located at /share/data/exomes/C1234.vcf.gz and DATA_DIR=/share/data. Then, we can use the following command to upload the variants. This annotation procedure does support uploading multi-sample vcf files.

chr Tags

EVAdb expects all contigs to have the chr Prefix. Make sure your contigs are named chr1-chrY prior to executing this command.

A simple command to change contig names would be the following (although there are some caveats when changing the reference like that). Set $f to your .vcf.gz input and $VCF to the desired output location.

zcat $f | awk '{ if ( $1 ~ "#" ) { print $0 } else { print "chr"$0 } }' > $VCF

docker-compose run annotation -vcf /data/exomes/C1234.vcf.gz \
  -sample C1234 \
  -se hg19_plus

Depending on the speed of your system the upload of one sample should be done in less than ten minutes. Afterwards you can try to filter your variant dataset using the tools in the EVAdb user application frontend.

Finally, fill the frequency tables with information, s.t. variant filters can work with the in-house data.

docker-compose run annotation bash
/pipeline/doAfterImport.pl -se hg19_plus -s 40