Import variation data

Variants can imported using a wrapper around the Ensembl Import VCF Script, which exposes a subset of the full functionality.

The input must be a bgzipped vcf file:

$ bgzip variants.vcf

Create and edit a panel file to associate samples with populations:

• this file has 2, tab-separated columns

• sample names must match the sample names in the vcf file

• only samples listed in this file will be imported

$ nano nano /path/to/data/panel.tsv
sample_1           population 1
sample_2           population 1
sample_3           population 2
sample_4           population 2
...

Create and edit a description file to add a description for each sample:

• this file has 2, tab-separated columns

• sample names must match the sample names in the vcf file

• html markup is supported and can be used to add a link to related SRA accession, if available

$ nano nano /path/to/data/description.tsv
sample_1           description of sample 1
sample_2           description of sample 2
sample_3           description of sample 3
sample_4           description of sample 4
...

Create and edit a configuration file to set database and variant details:

• as with the core database import, common settings can be specified in a default.ini file and passwords can be set in an overwrite.ini file

• if database connection settings are not set in a [DATABASE_VARIATION] section, values from [DATABASE_CORE] will be reused

• The variation database name must match the corresponding core database name with "variation" in place of "core"

• when importing local files, specify the path to the file as mounted in the container

• the FILTER will be passed to the bcftools view command with the -i flag, this is not needed if your SNP data are already filtered

e85

$ nano nano /path/to/conf/example_variants.ini
[DATABASE_CORE]
    NAME = heliconius_erato_demophoon_v1_core_32_85_1
    HOST = genomehubs-mysql
    PORT = 3306
    RW_USER = importer
    RW_PASS = CHANGEME
    RO_USER = anonymous
[DATABASE_VARIATION]
    NAME = genus_species_assembly_variation_32_85_1
[META]
    SPECIES.PRODUCTION_NAME = genus_species_assembly
    SPECIES.SCIENTIFIC_NAME = Genus species
[FILES]
    VCF = [ vcf /import/data/variants.vcf.gz ]
    PANEL = [ tsv /import/data/panel.tsv ]
[STUDY]
    SOURCE = Anonymous 2017
[BCFTOOLS]
    FILTER = QUAL>=30 & FMT/DP>=10 & FMT/DP<=100 & SUM(FMT/DP)<=N_SAMPLES*100 & FMT/SB<200 & MIN(FMT/GQ)>=30

e89

Run the GenomeHubs variation container:

• depending on the number of SNPs in your VCF file after filtering, is likely to take several hours to run

e85

docker run --rm \
           -d \
           --name genomehubs-variation \
           -u $UID:$GROUPS \
           -v /path/to/conf:/import/conf \
           -v /path/to/data:/import/data \
           -e FLAGS="-i" \
           -e VARIANTS=example_variants \
           genomehubs/variation:17.03

e89

Modify the EasyMirror configuration to load variation databases:

• EasyMirror will attempt to load database types listed in SPECIES_DB_AUTOEXPAND so this can be used to load funcgen, etc databases mirrored from Ensembl

$ nano ~/genomehubs/v1/ensembl/conf/setup.ini
[DATA_SOURCE]
  SPECIES_DB_AUTOEXPAND = [ variation ]

Restart your Ensembl site to load the newly created variation database:

e85

$ docker rm -f genomehubs-ensembl
$ docker run -d \
             --name genomehubs-ensembl \
             -v ~/genomehubs/v1/ensembl/conf:/ensembl/conf:ro \
             --link genomehubs-mysql \
             -p 8081:8080 \
             genomehubs/easy-mirror:17.03

e89