Setup and Customize

Use environment variables


store the path to transvar.cfg

export TRANSVAR_CFG=path_to_transvar.cfg

If not specified, TransVar will use [installdir]/lib/transvar/transvar.cfg directory or your local ~/.transvar.cfg if the installation directory is inaccessible.


store the path to the directory where auto-download of annotation and reference go

export TRANSVAR_DOWNLOAD_DIR=path_to_transvar_download_directory

If not specified, TransVar will use [installdir]/lib/transvar/ directory or your local ~/ if the installation directory is inaccessible.

Install and specify reference genome assembly

Download from TransVar database

For some genome assembly (currently hg18, hg19, hg38, mm9 and mm10) we provide download via

transvar config --download_ref --refversion [reference name]

See transvar config -h for all choices of [reference name]).

Manual download and index

For other genome assemblies, one could manually download the genome as one file and index it manually by,

samtools faidx [fasta]

Once downloaded and indexed, the genome can be used through the “–reference” option followed by path to the genome:

transvar ganno -i "chr1:g.30000000_30000001" --gencode --reference path_to_hg19.fa

or “–refversion” followed by the short version id.

transvar ganno -i "chr1:g.30000000_30000001" --gencode --refversion hg19

One can store the location in transvar.cfg file. To set the default location of genome file for a reference version, say, to path_to_hg19.fa,

transvar config -k reference -v path_to_hg19.fa --refversion hg19

will create in transvar.cfg an entry

reference = hg19.fa

so that there is no need to specify the location of reference on subsequent usages.

Install and specify transcript annotations

Download from TransVar database

One could automatically download transcript annotations via E.g.,

transvar config --download_anno --refversion hg19

which download annotation from TransVar database to [installdir]/lib/transvar/ directory or your local ~/ if the installation directory is inaccessible. See transvar config -h for all version names. These will also create default mappings under the corresponding reference version section of transvar.cfg like

ucsc = /home/wzhou1/download/hg19.ucsc.txt.gz

Index from GTF files

TransVar databases can be obtained from indexing a GTF file. For example,

transvar index --refseq hg38.refseq.gff.gz

The above will create a bunch of transvar databaase files with the suffix hg38.refseq.gff.gz.transvardb*.

Download from Ensembl ftp

One also has the option of downloading from Ensembl collection.

transvar config --download_ensembl --refversion mus_musculus

Without specifying the refversion, user will be prompted a collection of options to choose from.

Know Current configuration

To show the location and the content of currently used transvar.cfg, one may also run

transvar config

which returns information about the setup regarding to the current reference selection, including the location of the reference file and database file.

Current reference version: mm10
reference: /home/wzhou/genomes_link/mm10/mm10.fa
Available databases:
refseq: /home/wzhou/tools/transvar/transvar/
ccds: /home/wzhou/tools/transvar/transvar/
ensembl: /home/wzhou/tools/transvar/transvar/

specifying --refversion displays the information under that reference version (without changing the default reference version setup).

Set default reference builds

To switch reference build

transvar config --switch_build mm10

switches the default reference build to mm10. This is equivalent to

transvar config -k refversion -v mm10

which sets the refversion slot explicitly.

Use Additional Resources

TransVar uses optional additional resources for annotation.


For example, one could annotate SNP with dbSNP id by downloading the dbSNP files. This can be done by

transvar config --download_dbsnp

TransVar automatically download dbSNP file which correspoding to the current default reference version (as set in transvar.cfg). This also sets the entry in transvar.cfg. With dbSNP file downloaded, TransVar automatically looks for dbSNP id when performing annotation.

transvar panno -i 'A1CF:p.A309A' --ccds
A1CF:p.A309A CCDS7243 (protein_coding)       A1CF    -
   chr10:g.52576004T>G/c.927A>C/p.A309A      inside_[cds_in_exon_7]

Note that in order to use dbSNP, one must download the dbSNP database through

transvar config --download_dbsnp

or by configure the dbsnp slot in the configure file via

transvar config -k dbsnp -v [path to dbSNP VCF]

Manually set path for dbSNP file must have the file tabix indexed.

Control the length of reference sequence

TransVar reduces the reference sequence in a deletion to its length when the deleted reference sequence is too long. For example

$ transvar ganno -i 'chr14:g.101347000_101347023del' --ensembl


chr14:g.101347000_101347023del       ENST00000534062 (protein_coding)        RTL1    -
   chr14:g.101347000_101347023del24/c.4074+29_4074+52del24/. inside_[3-UTR;noncoding_exon_1]

where the deletion sequence is reduced to its length (del24). The –seqmax option changes the length threshold (default:10) when this behavior occur. When –seqmax is negative, the threshold is lifted such that the reference sequence is always reported regardless of its length, i.e.,

$ transvar ganno -i 'chr14:g.101347000_101347023del' --ensembl --seqmax -1

outputs the full reference sequence:

chr14:g.101347000_101347023del       ENST00000534062 (protein_coding)        RTL1    -
   chr14:g.101347000_101347023delTTGGGGTGAGAAATAGAGGGGACT/c.4074+29_4074+52delAGTCCCCTCTATTTCTCACCCCAA/.     inside_[3-UTR;noncoding_exon_1]