FAQ about submitting data to ClinVar

Getting started

1. Can I submit by API?
2. Which spreadsheet should I use?
3. Do I have to fill out all the tabs?
4. The spreadsheet template has a lot of columns; I don't know where to start.
5. I'm submitting my research on a clinically relevant variant for publication. Can I submit the data to ClinVar?
6. There is already a record in ClinVar for a variant that I intended to submit. Should I submit it too?
7. Can computational predictions of pathogenicity be submitted to ClinVar?

Getting into the details

1. I am not certain how to report disease or phenotype. Where do I begin?
2. Our clinical lab uses data from our own testing and literature curation to determine pathogenicity of variants. Can I submit both kinds of data?
3. I am submitting to ClinVar a variant that has been published. How do I make it clear that the publication describes a case that was tested in my clinical lab, as opposed to a paper that I curated from the literature?
4. Can I submit allele frequency data?
5. I'm submitting a large deletion, confirmed by sequencing across the breakpoint. Can I include that sequence in my submission?
6. How does ClinVar know if my submission is novel or an update to what I submitted previously?
7. How do I delete a record that I've submitted to ClinVar?
8. Which resources can I reference in my ClinVar submission?
9. Can I submit a genotype?

Getting started

Can I submit by API?

Yes! Please read our documentation for the ClinVar submission API.

Which spreadsheet should I use?

Most submitters will find the full spreadsheet, SubmissionTemplate, to be appropriate. There are a small number of required fields; after that, you can provide as many other data types as you would like to submit, such as the number of times a variant was observed, zygosity, observed phenotype, or family history.

If you are providing minimal data, you can use SubmissionTemplateLite; it has the same required fields as the full template but with fewer options for additional data. Keep in mind that, for ClinVar users, submissions with more supporting evidence are more useful.

Do I have to fill out all the tabs?

All submitters need to fill out the Variant tab. This tab should include each unique variant-condition pair and the assertion that you are making about them, one per row. If you are using SubmissionTemplate, aggregate-level supporting evidence, such as the allele origin or the number of times the variant was observed, is submitted on the Variant tab also.

If you are using SubmissionTemplateLite, you also need to fill out the ExpEvidence tab. This tab includes the aggregate-level supporting evidence for each assertion that you are making. For example, you can provide total counts for the number of times a variant has been observed, or you can provide aggregated counts broken down by variables such as sex, age range, or affected status.

If you are using SubmissionTemplate, there is the option to use the CaseData tab to provide information about each individual observed with the variant. The information is provided on CaseData as one individual per row. For each variant-condition assertion, please provide *either* aggregate data on the Variant tab or individual data on the CaseData tab, to avoid over-counting individuals.

The FunctionalEvidence tab is optional and is intended for research laboratories who are submitting functional data to support assertions of pathogenicity, such as animal models or in vitro cellular assays.

The spreadsheet template has a lot of columns; I don't know where to start.

If you are submitting a single variant interpretation, consider using the ClinVar submission wizard which guides you through the submission.

If you are submitting more than one variant on the spreadsheet template, consider starting with the required columns, as there are a small number of these. After that, consider what data you have readily available and how that relates to fields requested by ClinVar. If you need assistance mapping your data fields to ClinVar's, please contact us at clinvar@ncbi.nlm.nih.gov. Also note what data types you find useful as a ClinVar user, and consider if you are able to provide that data for your own submission.

I'm submitting my research on a clinically relevant variant for publication. Can I submit the data to ClinVar?

Yes, ClinVar welcomes submissions from research labs as well as from clinical testing labs. Submissions from research labs should use the collection method "research". We can hold your submission until published (see our hold until published policy and we can provide ClinVar SCV accession numbers that can be included in your publication.

There is already a record in ClinVar for a variant that I intended to submit. Should I submit my data for that variant too?

That depends on whether you are submitting variant-level or individual-level data. If you are submitting your own interpretation of a variant that is already represented in the database, then ClinVar welcomes your submission. Each additional submission adds value to the community curation of that variant by noting additional diagnostic terms, methods of assessment, and numbers of observations. However, if you are submitting research or clinical data for an individual that you know is already in ClinVar as a research or clinical report (i.e. not a literature report), then we ask that you do not re-submit that data to prevent the individual from being counted as more than one independent observation of the variant.

Can computational predictions of pathogenicity be submitted to ClinVar?

Variants that are classified based solely on computational predictions are not appropriate for ClinVar. Submissions that use software for pathogenicity prediction as part of the overall assessment are welcome; we encourage you to indicate the name of the software in your submission as part of the methods.

Getting into the details

I am not certain how to report disease or phenotype. Where do I begin?

We strongly recommend you submit using disease or phenotype information based on identifiers from groups that provide concepts, their definitions, and stable identifiers for those concepts.  Staff of ClinVar, GTR, and MedGen maintain this tutorial to aid your evaluation.

ClinVar provides options to report several classes of phenotypic information. These options are offered to conform to the types of information our submitters have when submitting their characterization of a variant.

If you are trying to select which term to use, please consider the following options:

Our clinical lab uses data from our own testing and literature curation to determine pathogenicity of variants. Can I submit both kinds of data?

Yes! Clinical testing data and data from literature review should be submitted on separate rows, whether you submit aggregate or individual data. Use the "Collection method" column to indicate whether the data are from "clinical testing" or "literature only". The distinction is important, because ClinVar aggregates cases reported from clinical testing but not from the literature, to avoid redundantly reporting the same case.

I’m submitting to ClinVar a variant that has been published. How do I make it clear that the publication describes a case that was tested in my clinical lab, as opposed to a paper that I curated from the literature?

Include the PubMed ID as a citation for your submission, which can have method type "clinical testing". Papers that describe the variant but not the submitted case can be submitted with method type "literature only" (see above), or the PubMed ID can be included in your "clinical testing" submission as a citation.

Can I submit allele frequency data?

ClinVar imports allele frequency data from the 1000 Genomes Project and from GO-ESP, so these data do not need to be included in your submission. Frequency data specific to a population that you are testing may be submitted.

I'm submitting a large deletion, confirmed by sequencing across the breakpoint. Can I include that sequence in my submission?

The breakpoint sequence may be submitted to GenBank, and the interpretation of the variant may be submitted to ClinVar.  This makes it easier for users to find the sequence data, and allows ClinVar to point to that sequence as supporting evidence for the assertion.

How does ClinVar know if my submission is novel or an update to what I submitted previously?

If you have already submitted information to ClinVar, we encourage you to submit an update if you have novel information.  If you have already submitted, you received an accession starting with SCV that uniquely identifies each variation-condition relationship that you submitted. The most reliable method for ClinVar to identify an update is for you to include the SCV accessions in your subsequent submissions and indicate that the variant is an update in the "Novel/update" column on the Variant tab. If the SCV accessions are not included, ClinVar will use the localID to match the submission to existing data in the database. If there is a match, it will be processed as an update; if no match, it will be processed as a novel record.  Updates may be global and may include novel, updated and unchanged variants from your dataset. In a global update, please include the SCV accessions for updated or unchanged variants, indicating "update" in the "Novel/update" column; for unchanged variants, please also include all the that was data originally submitted. For novel variants, no accession is necessary unless you reserved accessions in advance.

How do I delete a record that I've submitted to ClinVar?

If you just need to delete the record, please use the "Deletes" worksheet on the submission spreadsheet template. In the "ClinVar Accession" column, provide the SCV accession numbers for the records to delete. You have the option of providing an explanation for deleting the record in the "Comment" column. If you do not provide a comment, a generic reason will be provided: "This record was deleted at the request of the submitter." Note that you can only delete your own SCV records; you cannot delete SCV records owned by another submitter.

If you want to delete the record but replace it with another of your own SCV records, resubmit the data for the SCV record that you wish to retain, including its SCV accession in the "ClinVar Accession" column. Select "update" in the "Novel or update" column, and include the SCV accession for the record you are not retaining in the "Replaces ClinVarAccession" column.

Which resources can I reference in my ClinVar submission?

Submitters may reference any public resource that does not restrict access, e.g. those that do not require a subscription.

Can I submit a genotype?

You can submit a genotype to ClinVar, but in most cases this is not appropriate. ClinVar is a database of variant-level interpretations, not patient-level interpretations. So the majority of interpretations in ClinVar are for individual variants, even if a variant was observed in combination with another variant.

Compound heterozygotes

In general, if you interpreted two variants in the same gene of an individual, the two variants should be submitted individually to ClinVar, each variant with its own interpretation. This makes it easier for others to find data in ClinVar for each individual variant, and it makes the interpretation for each variant clear. An interpretation for the compound heterozygote is acceptable if the combination of the two variants is important for the interpretation, e.g. typically each variant is benign but the combination is pathogenic.

Note that in ClinVar, a "compound heterozygote" refers to an individual with two variants in trans in the same gene, where both of the variants are considered to be pathogenic or likely pathogenic. Data for a heterozygote with two variants in trans that are of uncertain significance should be submitted as separate records, one for each variant.

Homozygotes

If you interpreted a variant that was observed in homozygosity, the variant should be submitted as a single variant, not as a genotype with the variant in homozygosity. You can indicate in the observation data that the variant was observed in homozygosity. Mode of inheritance can also be provided, e.g. submit "autosomal recessive inheritance" to indicate that while the variant is interpreted as pathogenic, a single copy of the variant does not cause disease.