ANNOVAR main package

You can post questions through Disqus in this website or just email me directly.

The latest version of ANNOVAR can always be downloaded here (registration required).

ANNOVAR is written in Perl and can be run as a standalone application on diverse hardware systems where standard Perl modules are installed.

Additional databases

Many of the databases that ANNOVAR uses can be directly retrieved from UCSC Genome Browser Annotation Database by -downdb argument.

Several very commonly used annotation databases for human genomes are additionally provided below. In general, users can use -downdb -webfrom annovar in ANNOVAR directly to download these databases. To view of full list of databases (and their size and last changed date) prepared by ANNOVAR developers, use avdblist keyword in -downdb operation.

NOTE: several whole-genome databases (cadd, cadd13, fathmm, dann, eigen, gerp++, etc.) are available for download after server migration in June 2019. Please do NOT download unless you absolutely need to use them in your whole genome analysis (note that each file is ~200GB in your local computer), since each download will cost me a few dollars now.

- For gene-based annotation

Build Table Name Explanation Date
hg18 refGene FASTA sequences for all annotated transcripts in RefSeq Gene (last update was 2020-08-22 at UCSC) 20211019
hg19 refGene same as above (last update was 2020-08-17 at UCSC) 20211019
hg38 refGene same as above (last update was 2020-08-17 at UCSC) 20211019
hg18 refGeneWithVer FASTA sequences for all annotated transcripts in RefSeq Gene with version number (last update was 2020-08-22 at UCSC) 20211019
hg19 refGeneWithVer same as above (the last update was 2020-08-17 at UCSC) 20211019
hg38 refGeneWithVer same as above (last update was 2020-08-17 at UCSC) 20211019
hg18 knownGene FASTA sequences for all annotated transcripts in UCSC Known Gene (last update was 2009-05-10 at UCSC) 20211019
hg19 knownGene same as above (last update was 2013-06-30 at UCSC) 20211019
hg38 knownGene same as above (last update was 2022-05-15 at UCSC) 20211019
hg18 ensGene FASTA sequences for all annotated transcripts in wgEncodeGencodeManualV3 collection (last update was 2010-02-28 at UCSC) 20211019
hg19 ensGene FASTA sequences for all annotated transcripts in Gencode v43 Basic collection lifted up to hg19 (last update was 2023-02-15 at UCSC) 20230315
hg38 ensGene FASTA sequences for all annotated transcripts in Gencode v43 Basic collection (last update was 2023-02-15 at UCSC) 20230315
hg19 ensGene40 FASTA sequences for all annotated transcripts in Gencode v40 Basic collection lifted up to hg19 (last update was 2022-04-28 at UCSC) 20220802
hg38 ensGene40 FASTA sequences for all annotated transcripts in Gencode v40 Basic collection (last update was 2022-04-21 at UCSC) 20220802
hg19 ensGene41 FASTA sequences for all annotated transcripts in Gencode v41 Basic collection lifted up to hg19 (last update was 2022-07-12 at UCSC) 20221005
hg38 ensGene41 FASTA sequences for all annotated transcripts in Gencode v41 Basic collection (last update was 2022-07-12 at UCSC) 20221005
hs1 refGene FASTA sequences for all annotated transcripts in RefSeq Gene in CHM13-T2T reference genome (last update was 2023-05-29 at UCSC) 20230811
---

- For filter-based annotation

Build Table Name Explanation Date
hg18 avsift whole-exome SIFT scores for non-synonymous variants (obselete and should not be uesd any more) 20120222
hg19 avsift same as above 20120222
hg18 ljb26_all whole-exome SIFT, PolyPhen2 HDIV, PolyPhen2 HVAR, LRT, MutationTaster, MutationAssessor, FATHMM, MetaSVM, MetaLR, VEST, CADD, GERP++, PhyloP and SiPhy scores from dbNSFP version 2.6 20140925
hg19 ljb26_all same as above 20140925
hg38 ljb26_all same as above 20150520
hg18 dbnsfp30a whole-exome SIFT, PolyPhen2 HDIV, PolyPhen2 HVAR, LRT, MutationTaster, MutationAssessor, FATHMM, MetaSVM, MetaLR, VEST, CADD, GERP++, DANN, fitCons, PhyloP and SiPhy scores from dbNSFP version 3.0a 20151015
hg19 dbnsfp30a same as above 20151015
hg38 dbnsfp30a same as above 20151015
hg19 dbnsfp31a_interpro protein domain for variants 20151219
hg38 dbnsfp31a_interpro same as above 20151219
hg18 dbnsfp33a whole-exome SIFT, PolyPhen2 HDIV, PolyPhen2 HVAR, LRT, MutationTaster, MutationAssessor, FATHMM, PROVEAN, MetaSVM, MetaLR, VEST, M-CAP, CADD, GERP++, DANN, fathmm-MKL, Eigen, GenoCanyon, fitCons, PhyloP and SiPhy scores from dbNSFP version 3.3a 20170221
hg19 dbnsfp33a same as above 20170221
hg38 dbnsfp33a same as above 20170221
hg18 dbnsfp35a same as above 20180921
hg19 dbnsfp35a same as above 20180921
hg38 dbnsfp35a same as above 20180921
hg18 dbnsfp35c same as above, suitable for commercial use 20181023
hg19 dbnsfp35c same as above 20181023
hg38 dbnsfp35c same as above 20181023
hg19 dbnsfp41a same as above 20201007
hg38 dbnsfp41a same as above 20201007
hg19 dbnsfp41c same as above, suitable for commerical use 20201007
hg38 dbnsfp41c same as above 20201007
hg18 dbnsfp42a reformatted to include more columns than dbnsfp41a 20210710
hg19 dbnsfp42a same as above 20210710
hg38 dbnsfp42a same as above 20210710
hg18 dbnsfp42c reformatted to include more columns than dbnsfp41a, commerical use 20210710
hg19 dbnsfp42c same as above 20210710
hg38 dbnsfp42c same as above 20210710
hg19 dbscsnv11 dbscSNV version 1.1 for splice site prediction by AdaBoost and Random Forest 20151218
hg38 dbscsnv11 same as above 20151218
hg19 intervar_20170202 InterVar: clinical interpretation of missense variants (indels not supported) 20170202
hg19 intervar_20180118 InterVar: clinical interpretation of missense variants (indels not supported) 20180325
hg38 intervar_20180118 InterVar: clinical interpretation of missense variants (indels not supported) 20180325
hg18 cg46 alternative allele frequency in 46 unrelated human subjects sequenced by Complete Genomics 20120222
hg19 cg46 same as above index updated 2012Feb22
hg18 cg69 allele frequency in 69 human subjects sequenced by Complete Genomics 20120222
hg19 cg69 same as above 20120222
hg19 cosmic64 COSMIC database version 64 20130520
hg19 cosmic65 COSMIC database version 65 20130706
hg19 cosmic67 COSMIC database version 67 20131117
hg19 cosmic67wgs COSMIC database version 67 on WGS data 20131117
hg19 cosmic68 COSMIC database version 68 20140224
hg19 cosmic68wgs COSMIC database version 68 on WGS data 20140224
hg19 cosmic70 same as above 20140911
hg18 cosmic70 same as above 20150428
hg38 cosmic70 same as above 20150428
hg19/hg38 cosmic71, 72, ..., 80 read here
hg18 esp6500siv2_ea alternative allele frequency in European American subjects in the NHLBI-ESP project with 6500 exomes, including the indel calls and the chrY calls. This is lifted over from hg19 by myself 20141222
hg19 esp6500siv2_ea same as above 20141222
hg38 esp6500siv2_ea same as above, lifted over from hg19 by myself 20141222
hg18 esp6500siv2_aa alternative allele frequency in African American subjects in the NHLBI-ESP project with 6500 exomes, including the indel calls and the chrY calls. This is lifted over from hg19 by myself. 20141222
hg19 esp6500siv2_aa same as above 20141222
hg38 esp6500siv2_aa same as above, lifted over from hg19 by myself 20141222
hg18 esp6500siv2_all alternative allele frequency in All subjects in the NHLBI-ESP project with 6500 exomes, including the indel calls and the chrY calls. This is lifted over from hg19 by myself. 20141222
hg19 esp6500siv2_all same as above 20141222
hg38 esp6500siv2_all same as above, lifted over from hg19 by myself 20141222
hg19 exac03 ExAC 65000 exome allele frequency data for ALL, AFR (African), AMR (Admixed American), EAS (East Asian), FIN (Finnish), NFE (Non-finnish European), OTH (other), SAS (South Asian)). version 0.3. Left normalization done. 20151129
hg18 exac03 same as above 20151129
hg38 exac03 same as above 20151129
hg19 exac03nontcga ExAC on non-TCGA samples (updated header) 20160423
hg38 exac03nontcga same as above 20160423
hg19 exac03nonpsych ExAC on non-Psychiatric disease samples (updated header) 20160423
hg38 exac03nonpsych same as above 20160423
hg38 exac10 No difference as exac03 based on this; use exac03 instead X
hg19 gene4denovo201907 gene4denovo database 20191101
hg38 gene4denovo201907 gene4denovo database 20191101
hg19 gnomad_exome gnomAD exome collection (v2.0.1) 20170311
hg38 gnomad_exome gnomAD exome collection (v2.0.1) 20170311
hg19 gnomad_genome gnomAD genome collection (v2.0.1) 20170311
hg38 gnomad_genome gnomAD genome collection (v2.0.1) 20170311
hg19 gnomad211_exome gnomAD exome collection (v2.1.1), with "AF AF_popmax AF_male AF_female AF_raw AF_afr AF_sas AF_amr AF_eas AF_nfe AF_fin AF_asj AF_oth non_topmed_AF_popmax non_neuro_AF_popmax non_cancer_AF_popmax controls_AF_popmax" header 20190318
hg19 gnomad211_genome same as above 20190323
hg38 gnomad211_exome same as above 20190409
hg38 gnomad211_genome same as above 20190409
hg38 gnomad30_genome version 3.0 whole-genome data 20191104
hg38 gnomad312_genome version 3.1.2 whole-genome data 20221228
hg38 gnomad40_exome version 4.0 whole-exome data 20231127
hg38 gnomad40_genome version 4.0 whole-genome data 20231127
hs1 gnomad_genome whole-genome data from here 20230830
hg19 kaviar_20150923 170 million Known VARiants from 13K genomes and 64K exomes in 34 projects 20151203
hg38 kaviar_20150923 same as above 20151203
hg19 hrcr1 40 million variants from 32K samples in haplotype reference consortium 20151203
hg38 hrcr1 same as above 20151203
hg19 abraom 2.3 million Brazilian genomic variants 20181204
hg38 abraom liftOver from above 20181204
hg18 1000g (3 data sets) alternative allele frequency data in 1000 Genomes Project 20120222
hg18 1000g2010 (3 data sets) same as above 20120222
hg18 1000g2010jul (3 data sets) same as above 20120222
hg18 1000g2012apr I lifted over the latest 1000 Genomes Project data to hg18, to help researchers working with hg18 coordinates 20120820
hg19 1000g2010nov same as above 20120222
hg19 1000g2011may same as above 20120222
hg19 1000g2012feb same as above 20130308
hg18 1000g2012apr (5 data sets) This is done by liftOver of the hg19 data below. It contains alternative allele frequency data in 1000 Genomes Project for ALL, AMR (admixed american), EUR (european), ASN (asian), AFR (african) populations 20130508
hg19 1000g2012apr (5 data sets) alternative allele frequency data in 1000 Genomes Project for ALL, AMR (admixed american), EUR (european), ASN (asian), AFR (african) populations 20120525
hg19 1000g2014aug (6 data sets) alternative allele frequency data in 1000 Genomes Project for autosomes (ALL, AFR (African), AMR (Admixed American), EAS (East Asian), EUR (European), SAS (South Asian)). Based on 201408 collection v4 (based on 201305 alignment) 20140915
hg19 1000g2014sep (6 data sets) alternative allele frequency data in 1000 Genomes Project for autosomes (ALL, AFR (African), AMR (Admixed American), EAS (East Asian), EUR (European), SAS (South Asian)). Based on 201409 collection v5 (based on 201305 alignment) 20140925
hg19 1000g2014oct (6 data sets) alternative allele frequency data in 1000 Genomes Project for autosomes (ALL, AFR (African), AMR (Admixed American), EAS (East Asian), EUR (European), SAS (South Asian)). Based on 201409 collection v5 (based on 201305 alignment) but including chrX and chrY data finally! 20141216
hg18 1000g2014oct (6 data sets) same as above 20150428
hg38 1000g2014oct (6 data sets) same as above 20150424
hg19 1000g2015aug (6 data sets) The 1000G team fixed a bug in chrX frequency calculation. Based on 201508 collection v5b (based on 201305 alignment) 20150824
hg38 1000g2015aug (6 data sets) same as above 20150824
hg19 gme Great Middle East allele frequency including NWA (northwest Africa), NEA (northeast Africa), AP (Arabian peninsula), Israel, SD (Syrian desert), TP (Turkish peninsula) and CA (Central Asia) 20161024
hg38 gme same as above 20161024
hg19 mcap M-CAP scores for non-synonymous variants 20161104
hg38 mcap same as above 20161104
hg19 mcap13 [M-CAP scores v1.3] 20181203
hg19 revel REVEL scores for non-synonymous variants 20161205
hg38 revel same as above 20161205
hg18 snp128 dbSNP with ANNOVAR index files 20120222
hg18 snp129 same as above 20120222
hg19 snp129 liftover from hg18_snp129.txt 20120809
hg18 snp130 same as above 20120222
hg19 snp130 same as above 20120222
hg18 snp131 same as above 20120222
hg19 snp131 same as above 20120222
hg18 snp132 same as above 20120222
hg19 snp132 same as above 20120222
hg18 snp135 I lifted over SNP135 to hg18 20120820
hg19 snp135 same as above 20120222
hg19 snp137 same as above 20130109
hg18 snp138 I lifted over SNP138 to hg18 20140910
hg19 snp138 same as above file and index updated 20140910
hg19 avsnp138 dbSNP138 with allelic splitting and left-normalization 20141223
hg19 avsnp142 dbSNP142 with allelic splitting and left-normalization 20141228
hg19 avsnp144 dbSNP144 with allelic splitting and left-normalization (careful with bugs!) 20151102
hg38 avsnp144 same as above 20151102
hg19 avsnp147 dbSNP147 with allelic splitting and left-normalization 20160606
hg38 avsnp142 dbSNP142 with allelic splitting and left-normalization 20160106
hg38 avsnp144 dbSNP144 with allelic splitting and left-normalization 20151102
hg38 avsnp147 dbSNP147 with allelic splitting and left-normalization 20160606
hg19 avsnp150 dbSNP150 with allelic splitting and left-normalization 20170929
hg38 avsnp150 dbSNP150 with allelic splitting and left-normalization 20170929
hg18 snp128NonFlagged dbSNP with ANNOVAR index files, after removing those flagged SNPs (SNPs < 1% minor allele frequency (MAF) (or unknown), mapping only once to reference assembly, flagged in dbSnp as "clinically associated") 20120524
hg18 snp129NonFlagged same as above 20120524
hg18 snp130NonFlagged same as above 20120524
hg19 snp130NonFlagged same as above 20120524
hg18 snp131NonFlagged same as above 20120524
hg19 snp131NonFlagged same as above 20120524
hg18 snp132NonFlagged same as above 20120524
hg19 snp132NonFlagged same as above 20120524
hg19 snp135NonFlagged same as above 20120524
hg19 snp137NonFlagged same as above 20130109
hg19 snp138NonFlagged same as above 20140222
hg19 nci60 NCI-60 human tumor cell line panel exome sequencing allele frequency data 20130724
hg18 nci60 same as above 20150428
hg38 nci60 same as above 20150428
hg19 icgc21 International Cancer Genome Consortium version 21 20160622
hg19 icgc28 International Cancer Genome Consortium version 28 20210122
hg38 icgc28 International Cancer Genome Consortium version 28 20210122
hg19 clinvar_20131105 CLINVAR database with Variant Clinical Significance (unknown, untested, non-pathogenic, probable-non-pathogenic, probable-pathogenic, pathogenic, drug-response, histocompatibility, other) and Variant disease name 20140430
hg19 clinvar_20140211 same as above 20140430
hg19 clinvar_20140303 same as above 20140430
hg19 clinvar_20140702 same as above 20140712
hg38 clinvar_20140702 same as above 20140712
hg19 clinvar_20140902 same as above 20140911
hg38 clinvar_20140902 same as above 20140911
hg19 clinvar_20140929 same as above 20141002
hg19 clinvar_20150330 same as above but with variant normalization 20150413
hg38 clinvar_20150330 same as above but with variant normalization 20150413
hg19 clinvar_20150629 same as above but with variant normalization 20150724
hg38 clinvar_20150629 same as above but with variant normalization 20150724
hg19 clinvar_20151201 Clinvar version 20151201 with separate columns (CLINSIG CLNDBN CLNACC CLNDSDB CLNDSDBID) 20160303
hg38 clinvar_20151201 same as avove 20160303
hg19 clinvar_20160302 Clinvar version 20160302 with separate columns (CLINSIG CLNDBN CLNACC CLNDSDB CLNDSDBID) 20171003
hg38 clinvar_20160302 same as above (updated 20171003 to handle multi-allelic variants) 20171003
hg19 clinvar_20161128 Clinvar version 20161128 with separate columns (CLINSIG CLNDBN CLNACC CLNDSDB CLNDSDBID) 20171003
hg38 clinvar_20161128 same as above (updated 20170215 to add missing header line; 20171003 to handle multi-allelic variants) 20171003
hg19 clinvar_20170130 Clinvar version 20170130 with separate columns (CLINSIG CLNDBN CLNACC CLNDSDB CLNDSDBID) 20171003
hg38 clinvar_20170130 same as above 20171003
hg19 clinvar_20170501 Clinvar version 20170130 with separate columns (CLINSIG CLNDBN CLNACC CLNDSDB CLNDSDBID) 20171003
hg38 clinvar_20170501 same as above 20171003
hg19 clinvar_20170905 Clinvar version 20170905 with separate columns (CLINSIG CLNDBN CLNACC CLNDSDB CLNDSDBID) 20171003
hg38 clinvar_20170905 same as above 20171003
hg19 clinvar_20180603 Clinvar version 20180603 with separate columns (CLNALLELEID CLNDN CLNDISDB CLNREVSTAT CLNSIG) 20180708
hg38 clinvar_20180603 same as above 20180708
hg19 clinvar_20190305 Clinvar version 20190305 with separate columns (CLNALLELEID CLNDN CLNDISDB CLNREVSTAT CLNSIG) 20190311
hg38 clinvar_20190305 same as above 20190316
hg19 clinvar_20200316 Clinvar version 20200316 with separate columns (CLNALLELEID CLNDN CLNDISDB CLNREVSTAT CLNSIG) 20200401
hg38 clinvar_20200316 same as above 20200401
hg19 clinvar_20210123 Clinvar version 20210123 with separate columns (CLNALLELEID CLNDN CLNDISDB CLNREVSTAT CLNSIG) 20210202
hg38 clinvar_20210123 same as above 20210202
hg19 clinvar_20210501 Clinvar version 20210501 with separate columns (CLNALLELEID CLNDN CLNDISDB CLNREVSTAT CLNSIG) 20210507
hg38 clinvar_20210501 same as above 20210507
hg19 clinvar_20220320 Clinvar version 20220320 with separate columns (CLNALLELEID CLNDN CLNDISDB CLNREVSTAT CLNSIG) 20220330
hg38 clinvar_20220320 same as above 20220320
hg19 clinvar_20221231 Clinvar version 20221231 with separate columns (CLNALLELEID CLNDN CLNDISDB CLNREVSTAT CLNSIG) 20230105
hg38 clinvar_20221231 same as above 20230105
hg19 popfreq_max_20150413 A database containing the maximum allele frequency from 1000G, ESP6500, ExAC and CG46 20150413
hg19 popfreq_all_20150413 A database containing all allele frequency from 1000G, ESP6500, ExAC and CG46 20150413
hg19 mitimpact2 pathogenicity predictions of human mitochondrial missense variants (see here 20150520
hg19 mitimpact24 same as above with version 2.4 20160123
hg19 regsnpintron prioritize the disease-causing probability of intronic SNVs 20180920
hg38 regsnpintron lifeOver of above 20180922
hg18 gerp++elem conserved genomic regions by GERP++ 20140223
hg19 gerp++elem same as above 20140223
mm9 gerp++elem same as above 20140223
hg18 gerp++gt2 whole-genome GERP++ scores greater than 2 (RS score threshold of 2 provides high sensitivity while still strongly enriching for truly constrained sites. ) 20120621
hg19 gerp++gt2 same as above 20120621
hg19 caddgt20 with score>20 20160607
hg19 caddgt10 CADD with score>10 20160607
hg19 cadd CADD 20140223
hg19 cadd13 CADD version 1.3 20170123
hg19 cadd13gt10 CADD version 1.3 score>10 20170123
hg19 cadd13gt20 CADD version 1.3 score>20 20170123
hg19 caddindel removed 20150505
hg19 fathmm whole-genome FATHMM_coding and FATHMM_noncoding scores (noncoding and coding scores in the 2015 version was reversed) 20160315
hg19 gwava whole genome GWAVA_region_score and GWAVA_tss_score (GWAVA_unmatched_score has bug in file), see ref. 20150623
hg19 eigen whole-genome Eigen scores, see ref 20160330

User-contributed datasets

Several generous ANNOVAR users provide additional annotation datasets that may help other users. These datasets are described below:

  • MitImpact2: pathogenicity predictions of human mitochondrial missense variants. This is prepared as filter-based annotation format and users can directly download from ANNOVAR (see table above).
  • regsnpintron: regSNP-intron uses a machine learning algorithm to prioritize the disease-causing probability of intronic SNVs. The columns are "fpr (False positive rate), disease Disease category (B: benign [FPR > 0.1]; PD: Possibly Damaging [0.05 < FPR <= 0.1]; D: Damaging [FPR <= 0.05]), splicing_site Splicing site (on/off). Splicing sites are defined as -3 to +7 for donor sites, -13 to +1 for acceptor sites.". This is prepared as filter-based annotation format and users can directly download from ANNOVAR (see table above).
  • LoFtool score: gene loss-of-function score percentiles. The smaller the percentile, the most intolerant is the gene to functional variation. The file can be downloaded here. Manuscript in preparation (please contact Dr. Joao Fadista - joao.fadista@med.lu.se). The authors would like to thank the Exome Aggregation Consortium and the groups that provided exome variant data for comparison. A full list of contributing groups can be found at http://exac.broadinstitute.org/about.
  • RVIS-ESV score: RVIS score measures genetic intolerance of genes to functional mutations, as described in Petrovski et al. Original RVIS was constructed based on patterns of standing variation in 6503 samples. The authors have recently constructed scores based on the ~61,000 samples from ExAC. There is high correlation, but more resolution for many genes. The ExAC cohort implementation is what we consider RVIS (v2). It can be downloaded here.
  • GDI score: the gene damage index (GDI) is describing the accumulated mutational damage for each human gene in the general population, and shows that highly mutated/damaged genes are unlikely to be disease-causing and yet they generate a big proportion of false positive variants harbored in such genes. Therefore removing high GDI genes is a very effective way to remove confidently false positives from WES/WGS data. More details were given in this paper. The data set includes general damage prediction (low/medium/high) for different disease type (all, Mendelian, cancer, and PID) and can be downloaded from here.
  • TMC-SNPDB: SNP database from whole exome data of 62 normal samples derived from cancer patients of Indian origin, representing 114, 309 unique germline variants. Read the manuscript here. It is useful for exome sequencing studies on Indian populations and can be downloaded from here.
  • GenoNet Scores: cell-specific functional elements predicted by GenoNet organized by chromosomes in many cell types. You must use the specific link to download the files.

Third-party datasets

Several third-party researchers have provided additional annotation datasets that can be used by ANNOVAR directly. However, users need to agree to specific license terms set forth by the third parties:

  • SPIDEX: SPIDEX 1.0 - Deep Genomics : (Xiong et al, Science 2015) Machine-learning prediction on how genetic variants affect RNA splicing. This dataset can be downloaded here.

Third-party software tools

Customprodbj is a Java-based tool for customized protein database construction. It can build the database on a single or multiple VCF files on single or multiple individuals. It can be accessed at here. Command line example: java -jar customprodbj.jar -f input_variant_file_list.txt -d annovar_database/humandb/hg19_refGeneMrna.fa -r annovar_database/humandb/hg19_refGene.txt -t -o out/.

David Baux created a convenient script to automate the procedure of generating ClinVar databases in ANNOVAR. The scripts are available here.