When Should You Request Custom Marker Supplementation on a Human SNP Array?

Figure 1. Custom marker requests should begin with a biological coverage gap and end with an assay-feasibility decision.

You may need additional markers if the standard array misses a previously reported association, a population-specific variant, a replication marker, a functional allele, or a poorly tagged region that is central to the project. That does not mean every requested SNP should be added. A useful supplementation request connects each variant to a study objective, verifies the coordinate and allele definition, checks whether the backbone already assays or tags it, and leaves room for technical review.

The practical decision has three routes. Use the standard backbone when it already supports the primary analysis. Evaluate supplementation when a small, defensible set of direct markers would close important gaps. Choose a separate targeted method when the request is too large, too technically difficult, or better handled by sequencing. The Human 85K SNP Genotyping Array Service can be reviewed in this framework, but the final marker count and feasible custom content remain project-specific.

TL;DR

1. Request supplementation for defined gaps, not as a general attempt to make an array "more complete."
2. Confirm genome build, chromosome, position, reference and alternate alleles, strand, and supporting evidence for every requested variant.
3. Check whether a requested SNP is already present, duplicated under another identifier, or adequately tagged by nearby content.
4. Rank markers by scientific importance because probe design, sequence context, and available capacity can force tradeoffs.
5. Use targeted SNP genotyping or targeted resequencing when the marker list or discovery objective no longer fits an array add-on.

Six Triggers for a Supplementation Review

1. A reported locus must be measured directly

A replication study may require the exact lead SNP from a publication, consortium protocol, or previous cohort. An imputed proxy may not satisfy the study definition. Direct inclusion can also simplify cross-study harmonization when several cohorts must report the same locus. Before requesting it, confirm that the published identifier still maps to the expected coordinate and alleles on the project genome build.

2. The cohort contains underrepresented ancestry

Standard backbones reflect the discovery panels and populations used during array design. A variant that is common and informative in the intended cohort may be absent or weakly tagged if that population was not well represented. Population-specific add-on markers can improve local coverage or imputation, but only when discovery data and an appropriate reference panel justify the selection. The project should compare proposed additions with the planned genotype imputation and phasing workflow, not treat supplementation and imputation as independent decisions.

3. A candidate or functional variant drives the hypothesis

Some studies begin with a short list of coding variants, regulatory variants, expression quantitative trait loci, pharmacogenomic alleles, or experimentally supported candidates. If the inference depends on direct genotyping of those alleles, their absence is a real content gap. Evidence tier matters. A replicated functional variant and an exploratory database hit should not receive the same priority.

4. A genomic region is poorly covered

Marker count averaged across the genome can hide local deserts. Complex LD, structural complexity, segmental duplication, high sequence homology, or sparse backbone content may leave a region underrepresented. Review the manifest and use a defined coverage metric before adding markers. A linkage disequilibrium analysis can show whether existing SNPs tag the target alleles in the relevant population.

5. A study needs cross-platform continuity

Longitudinal programs may need a bridge between an older panel and a new backbone. Adding a limited set of legacy markers can preserve sample identity checks, historical comparisons, or meta-analysis compatibility. This is a version-control requirement rather than a genome-wide coverage argument. The team should document exactly which historical dataset, allele coding, and assay version must be matched.

6. Follow-up analysis requires a fixed shortlist

After GWAS, fine mapping, or functional prioritization, a project may need to genotype a defined shortlist in a larger replication cohort. The GWAS variant prioritization workflow illustrates why a lead signal is usually the beginning of locus interpretation rather than the final marker list. Supplementation is reasonable when the shortlist remains compatible with the array workflow. If many variants across broad intervals are needed, targeted sequencing deserves comparison.

Standard Backbone, Supplementation, or Another Method?

Figure 2. The right route depends on whether the project needs genome-wide context, direct fixed loci, or new variant discovery.

The distinction between direct assay and tagged coverage is important. A standard array can support an association through LD without directly typing the causal allele. Conversely, adding a biologically attractive SNP does not guarantee useful information if the assay fails or the variant is nearly monomorphic in the cohort. Review the broader tag-SNP selection principles when the purpose of an addition is regional coverage rather than direct replication.

Information Needed for Every Requested Marker

Do not send only a list of rsIDs. Identifiers can merge, be withdrawn, map differently between assemblies, or represent alleles on the opposite strand. A structured submission table makes feasibility review faster and reduces preventable allele errors.

Provide the table as a versioned file. Do not mix GRCh37 and GRCh38 coordinates in the same unlabelled column. If only rsIDs are available, state that explicitly and request normalization before design. For insertions, deletions, multi-allelic sites, and palindromic SNPs, include additional sequence context and the intended allele representation.

Technical Feasibility Is a Separate Gate

A scientifically valuable marker can still be unsuitable for the array chemistry. Probe design depends on the flanking sequence, nearby polymorphisms, repeats, paralogous sequence, GC composition, variant type, and platform-specific rules. Strand choice and allele conversion also require care. A request should therefore distinguish "must answer biologically" from "must be assayed by this exact marker." Sometimes a nearby proxy can serve the analytical need more reliably.

Feasibility review normally asks:

1. Is the variant biallelic and represented consistently in current databases?
2. Is adequate unique flanking sequence available for probe design?
3. Do nearby variants interfere with hybridization or allele discrimination?
4. Is the locus duplicated, repetitive, or located in a complex genomic region?
5. Does the requested allele orientation match the genome build and reporting convention?
6. Is there enough custom capacity after required backbone and control content?
7. Can a proxy marker meet the objective if the preferred variant fails design?

Do not interpret "submitted for design" as "guaranteed to work." Candidate content needs in-silico review, manufacturing assessment, clustering, and empirical performance evaluation. Recent custom-array studies have reported high overall conversion while still showing that some marker classes or genomic contexts perform less well. A representative pilot should therefore include samples that carry the relevant alleles whenever possible.

Prioritize Markers Before Capacity Becomes the Constraint

A flat marker list creates avoidable conflict at the design stage. Divide candidates into functional groups and assign a decision rule to each group.

1. Required direct markers: The project cannot meet its stated objective without the exact allele.
2. High-priority regional markers: Improve coverage of a validated locus or population-specific haplotype.
3. Replication and harmonization markers: Preserve compatibility with named studies, panels, or legacy datasets.
4. Exploratory candidates: Interesting but not essential to the primary endpoint.
5. Backup proxies: Preselected alternatives for technically difficult required markers.

Within each group, remove duplicates and check backbone overlap. Then evaluate allele frequency, LD redundancy, genome distribution, and evidence strength. This process prevents custom capacity from being consumed by many correlated SNPs while leaving another priority region uncovered.

When a Targeted Method Is Better

Figure 3. A growing marker list, difficult loci, or a discovery objective can shift the project away from array supplementation.

Supplementation is not the default answer to every gap. A SNP Genotyping Service may fit a compact, fixed set of known alleles when genome-wide backbone data are unnecessary. Targeted Resequencing is more appropriate when the goal is to discover variants across exons, genes, LD blocks, or regulatory intervals. Broader sequencing may be necessary when population-specific variation is poorly characterized. The existing comparison of SNP arrays, low-pass WGS, and deep WGS can help place those alternatives in context.

Consider leaving the array route when:

1. The project needs discovery rather than direct genotyping of known variants.
2. Requested content spans long or numerous genomic intervals.
3. Indels, structural variants, paralogous regions, or complex haplotypes dominate the list.
4. The custom list becomes large relative to the standard backbone.
5. The same focused marker set will be reused frequently without a genome-wide analysis need.
6. Many required variants fail probe feasibility and acceptable proxies are unavailable.

Platform choice should be based on the analyzable dataset required at the end, not on the wish to keep every assay on one technology.

Pilot and Acceptance Criteria

A pilot is useful only when pass criteria are defined before testing. Include diverse samples from the intended cohort, DNA inputs representative of production material, replicates, and reference samples with known genotypes if available. Evaluate both sample-level and marker-level performance.

Useful acceptance criteria include call rate, cluster separation, replicate concordance, agreement with orthogonal genotypes, Mendelian consistency for available trios, allele-frequency plausibility, Hardy-Weinberg review where appropriate, and missingness by population or processing group. A marker that passes globally but fails in the population carrying the research question should not be accepted without investigation.

The pilot should also test data integration. Confirm genome build, allele coding, variant identifiers, and merge behavior with legacy or consortium datasets. Custom markers need the same version control as the rest of the manifest. Record which submitted variants were accepted, rejected, substituted, or deferred, and retain the reason.

A Practical Submission Sequence

1. Define the biological and analytical reason for supplementation.
2. Compare the requested variants with the current array manifest and population LD.
3. Normalize identifiers, coordinates, alleles, and genome build.
4. Rank required markers, optional markers, and backup proxies.
5. Review in-silico probe feasibility and available custom capacity.
6. Compare the supplemented array with focused genotyping and sequencing alternatives.
7. Lock a versioned candidate manifest and run a representative pilot.
8. Apply predefined marker and sample acceptance criteria before production.

This sequence keeps the conversation centered on coverage gaps, not on an arbitrary add-on count. It also produces the files a laboratory needs to review feasibility and estimate the production workflow.

When Supplementation May Not Add Value

Consider a study that requests dozens of published index SNPs from several ancestry groups. Adding every index marker may consume capacity without improving the intended analysis: some variants may already be well tagged by the backbone, some associations may not transfer to the cohort, and others may be identifiers for a region rather than plausible functional variants. Classify each request as mandatory direct observation, preferred direct observation, proxy-eligible, or exploratory. Then review LD and allele frequency in the relevant population and keep a documented reason for inclusion or rejection.

When a requested marker fails technical review, first verify build, strand, alleles, and flanking sequence. Next assess nearby polymorphisms, repetitive sequence, paralogy, and whether a population-appropriate proxy exists. A proxy supported in one population should not be assumed to perform equivalently in another. If no reliable direct or proxy assay is available, report the gap rather than treating attempted inclusion as successful coverage. For a critical locus, orthogonal genotyping or targeted resequencing may be a more transparent choice.

Custom content can improve measurement of selected variants, but it does not make an association causal or ensure that untyped variation in the region is captured. Replication, fine mapping, and functional validation may still be needed. Published conversion rates or pilot thresholds from another array design are useful context, not universal guarantees. Project acceptance criteria should specify which markers are indispensable, how concordance will be assessed, what happens to systematic failures, and whether the final production manifest remains scientifically adequate after exclusions.

Frequently Asked Questions

For research purposes only. The information and services described here are not intended for clinical diagnosis, therapeutic decisions, or personal health assessment.

  Related Services

SNP Genotyping Service

GBS & Marker-Assisted Selection

Human 85K SNP Genotyping Array Service

Sequencing

Whole Exome Sequencing for Population Genetics

Bioinformatics

Genetic Diversity Analysis

Genome-wide Association Analysis Service

  Download

Power Your Regulatory Discoveries with DAP-seq: A MYB-GLK Case Study

EM-seq Services for High-Resolution DNA Methylation Analysis

Multi-Omics Strategies to Decode Molecular Mechanisms at Every Level

Explore Customized Bioinformatics Services: Case Studies & Insights