BYSTRO AI

Query Allele Frequency

Filter variants by population allele frequency using gnomAD data, the largest database of human genetic variation. Identify rare, common, or population-specific variants for your research.

About gnomAD

The Genome Aggregation Database (gnomAD) is the largest collection of human genetic variation data, containing allele frequencies from diverse global populations.

Learn more in the gnomAD overview.

Quick Filtering with MAF

The simplest way to filter by rarity is using minor allele frequency (MAF). This searches across both gnomAD genomes and exomes datasets simultaneously for comprehensive population frequency filtering.

Common MAF Thresholds

Ultra-raremaf < 0.001
Less than 0.1% frequency — strongest filter for Mendelian disease candidates
Raremaf < 0.01
Less than 1% frequency — standard threshold for rare variant analysis
Low frequencymaf < 0.05
Less than 5% frequency — captures low-frequency variants across populations

Example: Finding Rare Variants

Find rare variants (frequency < 1%) using maf < 0.01:

Image showing how to search by minor allele frequency with maf < 0.01

Using MAF to quickly identify rare variants across all gnomAD populations

MAF searches both datasets automatically
Using maf as a query automatically searches both gnomAD genomes and gnomAD exomes datasets, providing comprehensive frequency filtering without needing to specify individual datasets.

Population-Specific Searches

For more precise population genetics analysis, query specific gnomAD population datasets using exact field names. This lets you identify variants that are rare in one population but common in another.

Example: Non-Finnish European Population

Search for variants rare in the Non-Finnish European population using gnomAD.genomes.af_nfe < 0.01:

Image showing how to search gnomAD by specific population with gnomAD.genomes.af_nfe < 0.01

Filtering for variants rare in the Non-Finnish European population using specific gnomAD field names

Available gnomAD Populations

gnomAD Genomes

  • gnomAD.genomes.afAll populations
  • gnomAD.genomes.af_afrAfrican / African American
  • gnomAD.genomes.af_amrLatino / Admixed American
  • gnomAD.genomes.af_asjAshkenazi Jewish
  • gnomAD.genomes.af_easEast Asian
  • gnomAD.genomes.af_finFinnish
  • gnomAD.genomes.af_nfeNon-Finnish European
  • gnomAD.genomes.af_othOther

gnomAD Exomes

  • gnomAD.exomes.afAll populations
  • gnomAD.exomes.af_afrAfrican / African American
  • gnomAD.exomes.af_amrLatino / Admixed American
  • gnomAD.exomes.af_asjAshkenazi Jewish
  • gnomAD.exomes.af_easEast Asian
  • gnomAD.exomes.af_finFinnish
  • gnomAD.exomes.af_nfeNon-Finnish European
  • gnomAD.exomes.af_sasSouth Asian
Complete field reference
Bystro reports many gnomAD fields including all population-specific frequencies. See our annotation field descriptions for detailed information about all available fields.

Important Considerations

Allele Frequency Reporting

Bystro reports allele frequencies relative to the specific variant allele in your dataset, not all previously observed variants at that position (which is how dbSNP reports frequencies).

gnomAD IDs

gnomAD ID shows only one rs-number per variant. Learn how this can be used to create Set IDs for SKAT analysis in our FAQ section.

Practical Applications

Rare Disease Research

Combine ultra-rare frequency filtering with high CADD scores to identify potentially pathogenic variants:

maf < 0.001 AND cadd > 20

Population Genetics

Compare allele frequencies between populations to identify population-specific variants:

gnomAD.genomes.af_eas > 0.05 AND gnomAD.genomes.af_nfe < 0.01

Clinical Variant Filtering

Focus on clinically relevant rare variants in coding regions:

maf < 0.01 AND refSeq.exonicAlleleFunction:nonSynonymous
Best practices

Consider your study population: Use population-specific frequencies when studying specific ethnic groups.

Account for sample size: Check allele count (AC) and allele number (AN) fields for reliability.

Combine datasets: Use both exomes and genomes data for comprehensive frequency assessment.

Validate rare variants: Always validate ultra-rare variants with additional evidence.

Performance note
Dataset used in examples: 1000 Genomes Project (73,452,337 variants in 27,192 genes, queries typically complete in ~0.5 seconds).