How do I use the Variant Annotation Database for my project?

To utilize the Variant Annotation Database for your clinical research project, simply click on the “Annotate with the Variant Annotation Database” option in the Analysis Options window of SeqMan NGen when you set up your project.

The Variant Annotation database will automatically be used to annotate the variants identified in your data.

What supplemental variant data is available in the Variant Annotation Database?

DNASTAR’s Variant Annotation Database contains variant information using coordinates from GRCh37 (hg19) and GRCh38. Annotations include information about the frequency of the variant in the general population, in specific populations, and in publications, as well as information concerning the variant’s impact on functionality. The annotation information comes from several sources, including:

• Mastermind, the comprehensive database of genomic literature from Genomenon, providing access to NGS variant data from millions of PubMed publications.
• 1000 Genomes Project and the Exome Sequencing Project (ESP)
• dbNSFP v 4.1 which includes:

• Functional impact predictions from Sorting Intolerant from Tolerant (SIFT), MutationTaster, and the Likelihood Ratio Test (LRT).
• Evolutionary conservation scores from SiPhy, PhyloP, PhastCons and GERP++.
• Pathogenicity and clinical significance impact from ClinVar.
• Basic amino acid sequence information, Database of Single Nucleotide Polymorphism (dbSNP) IDs, and annotations from Uniprot and Interpro.

How can I find citations for a specific variant in current publications?

Our integration with Mastermind provides a seamless way to find publications on a specific variant of interest. In ArrayStar, easily access and link to data from Mastermind in the SNP Table for each of your variants, including:

• MM Count 1: Count of articles in the Mastermind Genomic Search Engine with cDNA matches for a specific variant
• MM Count 2: Count of articles in the Mastermind Genomic Search Engine with variants either explicitly matching at the cDNA level or given only at protein level.
• MM Count 3: Count of Mastermind articles in the Mastermind Genomic Search Engine, including other DNA-level variants resulting in the same amino acid change.
• MM Gene: Genes for a variant from the Mastermind Genomic Search Engine.
• MM HGVS: HGVS genomic notation for a variant from the Mastermind Genomic Search Engine.
• MM ID 3: Variant identifiers in the Mastermind Genomic Search Engine, as gene:key, for MMCNT3.
• MM URI 3: Search URI for articles in the Mastermind Genomic Search Engine, including other DNA-level variants resulting in the same amino acid change.

Is there an additional cost to access the Variant Annotation Database?

No. Access to the variant databases for human genome analysis is included with your purchase of Lasergene Genomics and remains available as long as your service plan is active.

Can the Variant Annotation Database be used with any Genome Template Package (GTP)?

No. The Variant Annotation Database is a human genome specific resource and must be used with either build 37 (GRCh37) or build 38 (GRCh38) of the human genome reference sequence. This is most easily done when setting up an assembly in SeqMan NGen where the Reference Sequence screen gives you the option to Download a Genome Package. The ensuing popup lets you choose GRCh37p.13 or GRCh38.p7, among others. (Note: When uploaded into SeqMan NGen, the Homo sapiens mitochondrion database will not work with the Variant Annotation Database).

Why do the 1000 Genomes minor allele frequencies (MAFs) reported in ArrayStar sometimes vary slightly between build 37 and build 38 based human genome assemblies?

In 2016, 1000 Genomes released their initial Phase 3 data set based on variant calls from 2,504 samples aligned to build 37 of the human genome (GRCh37). In 2018, an expanded data set was released of 2,548 samples aligned to build 38 of the human genome (GRCh38). 1000 Genomes has used “Phase 3” to describe both the 2,504 and 2,548 sample sets across both genome versions and four years of updates. DNASTAR is using the dataset version described in 1000 Genomes “Phase 3” callset for GRCh37 update.

The minor differences in the calculated MAFs between builds 37 and 38 are a reflection of the slight difference in total number of samples as well as potential differences in the number of samples with the specific allele being considered.

Why do the minor allele frequencies (MAFs) reported in ArrayStar sometimes vary from those on the 1000 Genomes and dbSNP websites?

As mentioned above, VAD-calculated 1000 Genome MAFs from GRCh37 and GRCh38 assemblies will differ slightly due to the slightly expanded number of samples used in the GRCh38 data set. dbSNP currently uses GRCh38 as their default reference but reports the 1000 Genomes MAFs based on the slightly smaller 2,504 sample data set. Therefore, VAD-calculated MAFs from GRCh37 based assemblies should be nearly identical with those at dbSNP except for very minor differences due to choices made for rounding. However, more noticeable differences may be seen between VAD-calculated MAFs from GRCh38-based assemblies and dbSNP due to the larger sample set used for the GRCh38 database. Reference: Announcement (2017) on Updated GRCh38 liftover.

Which versions of dbNSFP and 1000 Genomes versions are used in ArrayStar?

The versions being used are shown in ArrayStar’s Set Up Preprocessing wizard screen. As of 2021, ArrayStar is using Human genome GRCh38 from dbNSFP 4.1 and 1000 Genomes Phase 3.

Clinvar uses fourteen different terms for assigning clinical significance. Why can I only filter my data set in ArrayStar using ten of these terms?

In ArrayStar, variants can be filtered based on whether or not a particular clinvar_clnsig term is associated with that SNP. To filter based on clinvar_clnsig terms, choose Filter > Filter Selected or Filter > Filter All, then press the Choose SNP Criteria button and the Pathogenecity tab.

Clinvar uses fourteen different terms for assigning clinical significance, ten of which are available for filtering in ArrayStar. Four terms were omitted from the filter dialog because of their potential to lead to ambiguous or misleading results when filtering variant data. Clinvar terms that can be used in filtering are: Association, Benign, Drug response, Histocompatibility, Likely benign, Likely pathogenic, Pathogenic, Protective, Risk factor, Uncertain significance. The four ClinVar terms ommitted in ArrayStar’s filtering dialog are: Affects, Conflicting data from submitters, Other, Not provided.

In ArrayStar’s Variant Table, why does the Mastermind column always link to the Mastermind Genomics Search Engine’s build 37 pages?

As of early 2021, the Mastermind Genomics Search Engine uses only build 37 for their database and website. Mastermind does provide some additional data to ArrayStar via a large file remapped to build 38. ArrayStar then augments this with even richer build 38 data. We have no information regarding if and when Genomenon, the maker of the Mastermind Genomics Search Engine, intends to fully support build 38.