SNP Genotyping Explained: The Technology Behind DNA Testing

When you order a DNA test from Helixline or any other consumer genetics company, your saliva sample goes through a sophisticated molecular process that reads hundreds of thousands of specific positions in your genome. This technology is called SNP genotyping, and it is the foundation of virtually all consumer DNA testing today. But what exactly is a SNP? How does a DNA chip work? And what are the capabilities and limitations of this technology?

This article provides a comprehensive, technically accurate explanation of SNP genotyping - from the molecular biology of single nucleotide polymorphisms to the engineering of microarray chips, the data processing pipeline, and how genotyping compares to whole genome sequencing. Whether you are a curious consumer wanting to understand your DNA results or a student of genomics, this guide will give you a solid foundation.

What Are SNPs? The Building Blocks of Human Variation

A SNP (single nucleotide polymorphism, pronounced "snip") is the simplest form of genetic variation: a single-letter change in the DNA sequence at a specific position in the genome. The human genome is a sequence of approximately 3.2 billion nucleotide bases (A, T, C, and G). At most of these positions, every human on Earth has the same letter. But at certain positions, different people have different letters.

For example, at a particular position on chromosome 15, most people might have the nucleotide C (cytosine). But some individuals have a T (thymine) at that same position instead. This single-letter difference is a SNP. Each individual carries a specific combination of these variants - your personal set of SNPs is a major part of what makes you genetically unique.

SNPs by the Numbers

• Frequency in the genome: On average, there is 1 SNP approximately every 300 base pairs across the human genome. This means the genome contains roughly 10 million positions where common variation exists
• Common SNPs: Approximately 10 million SNPs have a minor allele frequency (MAF) greater than 1% in the global population - meaning at least 1 in 100 people carry the less common variant. These are the "common" SNPs
• Individual variation: Any given person typically carries 4-5 million SNPs that differ from the human reference genome (GRCh38). This means about 0.1-0.15% of your genome differs from the reference
• Shared variation: Any two unrelated humans differ at approximately 3-4 million SNP positions - about 99.9% of our DNA is identical
• Coding SNPs: Approximately 20,000-25,000 SNPs fall within the protein-coding regions of genes (exons), where they can directly affect protein structure and function
• The dbSNP database: The NCBI's comprehensive database of known variants contains over 600 million validated SNPs, including both common and rare variants

Why SNPs Matter

SNPs are not just random differences - many of them have functional consequences or serve as powerful markers for understanding human biology and history:

• Ancestry informative markers (AIMs): Certain SNPs show dramatically different frequencies across populations. For example, a SNP might be found in 85% of South Asians but only 5% of Europeans. By analyzing hundreds of thousands of such markers together, we can precisely determine an individual's ancestral origins. This is the basis of all DNA ancestry testing
• Disease associations: Genome-wide association studies (GWAS) have identified thousands of SNPs associated with disease risk. For example, specific SNPs in the HLA region on chromosome 6 are strongly associated with type 1 diabetes and celiac disease. SNPs near the BRCA1 gene region are associated with breast cancer risk
• Pharmacogenomics: SNPs in drug-metabolizing enzymes determine how individuals respond to medications. The CYP2D6 gene contains SNPs that affect metabolism of codeine, tamoxifen, and many antidepressants. The VKORC1 gene has SNPs that influence warfarin dosing
• Physical traits: SNPs determine visible traits like eye color (OCA2/HERC2 genes), hair color (MC1R gene), lactose tolerance (LCT gene), and skin pigmentation (SLC24A5, SLC45A2 genes)
• Evolutionary history: The distribution of SNPs across populations tells the story of human migration, natural selection, and adaptation. For example, the lactase persistence SNP (-13910*T) near the LCT gene tracks the spread of dairy farming cultures

How DNA Microarray Technology Works

The core technology behind SNP genotyping is the DNA microarray (also called a DNA chip, gene chip, or SNP array). It is an elegant piece of molecular engineering that allows hundreds of thousands of SNP positions to be read simultaneously in a single experiment.

The Physical Chip

A DNA microarray is a small glass or silicon slide - roughly the size of a postage stamp - onto which millions of microscopic spots of synthetic DNA have been deposited or grown in a precise grid pattern. Each spot contains billions of copies of a short, single-stranded DNA sequence called a probe. Each probe is designed to be complementary to the DNA sequence surrounding a specific SNP position in the human genome.

For a genotyping array that tests 700,000 SNPs, the chip contains at least 700,000 distinct probe spots (and typically more, with multiple probes per SNP for quality control). The probes are typically 25-50 nucleotides long and are arranged in a known pattern so that the location of each probe on the chip corresponds to a specific SNP.

The Genotyping Process: Step by Step

1. DNA Extraction: DNA is extracted from the saliva sample using chemical lysis (breaking open cells) and purification. A saliva sample typically yields 5-10 micrograms of high-quality genomic DNA, which is far more than needed for genotyping
2. DNA Amplification: The extracted DNA is amplified (copied) using whole-genome amplification (WGA), which creates thousands of copies of the entire genome. This ensures there is sufficient DNA material for the chip to work with. The amplified DNA is typically fragmented into pieces of 200-500 base pairs
3. Denaturation: The amplified DNA is heated to separate the double-stranded DNA into single strands. This is necessary because hybridization (the next step) requires single-stranded DNA
4. Hybridization: The single-stranded DNA fragments are washed over the microarray chip. Through the principle of complementary base pairing (A binds to T, C binds to G), each DNA fragment binds to the probe on the chip that matches its sequence. This process typically takes 16-24 hours at controlled temperature
5. Fluorescent Labeling and Detection: The bound DNA fragments are labeled with fluorescent molecules - typically two different colors corresponding to the two possible alleles at each SNP. For Illumina's technology (the most widely used), the two alleles at each SNP position generate different fluorescent signals:
   • Allele A might generate a red fluorescent signal
   • Allele B might generate a green fluorescent signal
   • A homozygous AA individual produces only red signal at that spot
   • A homozygous BB individual produces only green signal
   • A heterozygous AB individual produces both red and green signals
6. Optical Scanning: A high-resolution laser scanner reads the fluorescent intensity at each probe location on the chip. The scanner captures images at different wavelengths corresponding to each fluorescent color
7. Genotype Calling: Computational algorithms (such as Illumina's GenCall or the open-source GenTrain algorithm) analyze the fluorescent intensity data and assign a genotype (AA, AB, or BB) to each SNP position. The algorithm uses cluster analysis - comparing each individual's signal to the distribution of signals from a large reference panel - to make accurate calls

The Illumina Infinium Assay: Industry Standard

The most widely used genotyping technology in consumer DNA testing is Illumina's Infinium assay, used on their Global Screening Array (GSA). Key technical details:

• Bead-based technology: Instead of flat probe spots, Illumina uses 3-micron silica beads, each coated with hundreds of thousands of copies of a specific probe sequence. These beads are randomly distributed into microwells etched into a glass slide
• BeadChip format: The GSA BeadChip can process 24 samples simultaneously on a single slide, with each sample being genotyped at ~700,000 SNP positions
• Single-base extension: Illumina's Infinium II assay uses a clever single-base extension approach where the probe hybridizes immediately adjacent to the SNP position, and a single fluorescently-labeled nucleotide is enzymatically added to match the actual allele present in the sample
• Resolution: The iScan system reads each bead with spatial resolution sufficient to distinguish individual 3-micron beads, achieving call rates exceeding 99.5% in high-quality samples

The Illumina Global Screening Array (GSA)

The Illumina GSA is the current industry-standard genotyping platform used by the majority of consumer DNA testing companies worldwide, including Helixline. It deserves special attention because its design directly determines what your DNA test can and cannot tell you.

What the GSA Includes

• Core content (~700,000 SNPs): The backbone of the GSA includes carefully selected SNPs optimized for multi-ethnic use, covering common variants across diverse global populations including South Asian, East Asian, African, European, and admixed populations
• Ancestry-informative markers: Thousands of SNPs specifically chosen because they differentiate between ancestral populations, enabling precise ancestry composition analysis
• Clinically relevant variants: The GSA includes pharmacogenomic markers (CYP2D6, CYP2C19, VKORC1, DPYD, TPMT), known pathogenic variants (BRCA1/2, Lynch syndrome genes, CFTR), and carrier screening variants for hundreds of recessive conditions
• GWAS-validated markers: SNPs that have been reliably associated with traits and diseases in genome-wide association studies, enabling polygenic risk score calculation
• Mitochondrial DNA SNPs: Markers on the mitochondrial genome for maternal haplogroup assignment
• Y-chromosome SNPs: Markers on the Y chromosome (in males) for paternal haplogroup assignment
• Custom content: Companies like Helixline can add custom SNP content to the base GSA design, adding markers particularly relevant to their target populations (e.g., South Asian-specific ancestry markers, India-relevant pharmacogenomic variants)

How Many SNPs Do Different Companies Test?

Company / Platform	Approximate SNPs Tested	Array Platform	Notable Features
Helixline	~700,000	Illumina GSA	South Asian-optimized content; Indian ancestry markers
23andMe (v5)	~640,000	Custom Illumina GSA	Health reports; FDA-authorized genetic health risk reports
AncestryDNA	~700,000	Custom Illumina GSA	Largest reference database; ThruLines family matching
MyHeritage DNA	~700,000	Illumina GSA	International reach; family tree integration
Living DNA	~660,000	Custom Illumina GSA	Fine-scale British and European ancestry
Illumina GSA v3 (base)	~654,000	Illumina GSA v3	Multi-ethnic backbone; customizable

Genotyping vs. Whole Genome Sequencing: A Detailed Comparison

One of the most common questions about DNA testing is the difference between SNP genotyping (what consumer DNA tests do) and whole genome sequencing (WGS). These are fundamentally different technologies with different strengths, limitations, and costs:

Feature	SNP Genotyping (Microarray)	Whole Genome Sequencing (WGS)	Whole Exome Sequencing (WES)	Targeted Panel Sequencing
What it reads	~700,000 pre-selected SNP positions	All 3.2 billion base pairs	~30 million base pairs (protein-coding exons only)	Specific genes of interest (50-500 genes)
Typical cost	$50-200	$200-1,000	$200-500	$200-2,000
Turnaround time	2-4 weeks	4-8 weeks	4-8 weeks	2-6 weeks
Data output size	~10-50 MB	~100-200 GB (raw); 5-10 GB (processed)	~5-10 GB (raw); 500 MB (processed)	~1-5 GB (raw)
Common variant detection	Excellent (if variant is on the chip)	Excellent (all variants detected)	Good (coding regions only)	Excellent (within targeted regions)
Rare variant detection	Poor (most rare variants not on chip)	Excellent	Good (coding regions only)	Excellent (within targeted regions)
Structural variant detection	Limited	Good (with sufficient coverage)	Limited	Limited
Best for	Ancestry, common traits, carrier screening, pharmacogenomics	Comprehensive genetic analysis, research, rare disease diagnosis	Clinical rare disease diagnosis, gene discovery	Known disease testing, cancer panels, specific clinical questions

The Key Trade-Off: Breadth vs. Depth

Think of it this way: genotyping is like checking specific pages in a book, while sequencing is like reading the entire book cover to cover. Genotyping is faster and cheaper because it only looks at known, pre-selected positions. Sequencing is more comprehensive but more expensive and generates vastly more data that requires more sophisticated analysis.

For most consumer applications - ancestry analysis, common trait prediction, carrier screening, pharmacogenomics, and relative matching - genotyping provides more than sufficient information at a fraction of the cost. Sequencing becomes necessary when searching for rare or novel mutations, diagnosing unresolved genetic conditions, or conducting comprehensive research.

What SNP Genotyping CAN Detect

Despite testing only ~0.02% of the genome, SNP genotyping arrays are remarkably powerful for a wide range of applications:

Ancestry and Population Genetics

• Ancestral population composition: By comparing your SNP profile against reference panels from known populations, algorithms can estimate what percentage of your ancestry derives from different ancestral groups (e.g., 65% South Asian, 20% Central Asian Steppe, 15% Southeast Asian)
• Maternal haplogroup (mtDNA): SNPs on the mitochondrial genome assign your direct maternal lineage to a haplogroup with a known geographic and temporal origin
• Paternal haplogroup (Y-DNA): SNPs on the Y chromosome (males only) assign your direct paternal lineage to a haplogroup
• Relative matching: By comparing shared IBD (identical-by-descent) segments between two genotyped individuals, the technology can detect relatives from identical twins to 5th cousins
• Admixture dating: The length distribution of ancestry-specific DNA segments can estimate when mixing between different ancestral populations occurred

Health and Clinical Applications

• Known disease-risk SNPs: Variants with established associations to conditions like type 2 diabetes, coronary artery disease, age-related macular degeneration, and many others
• Carrier status: Heterozygous status for recessive disease alleles such as cystic fibrosis (CFTR gene), sickle cell disease (HBB gene), and beta-thalassemia
• Pharmacogenomics: Variants in drug-metabolizing enzymes that predict drug response (CYP2D6 poor/ultra-rapid metabolizer status, warfarin sensitivity, clopidogrel response)
• Polygenic risk scores: Combining the effects of thousands of individually small-effect SNPs to estimate overall genetic risk for complex diseases
• Runs of homozygosity (ROH): Detecting long stretches of homozygous DNA that indicate parental relatedness or population-level endogamy

Trait Prediction

• Physical traits: Eye color, hair color, skin pigmentation, earwax type (wet/dry), cilantro taste perception, asparagus smell detection
• Metabolic traits: Lactose tolerance/intolerance, caffeine metabolism speed, alcohol flush reaction, bitter taste sensitivity (TAS2R38 gene)
• Nutritional genetics: Vitamin metabolism variants, folate metabolism (MTHFR gene), vitamin D receptor variants

What SNP Genotyping CANNOT Detect

Understanding the limitations of genotyping is just as important as understanding its capabilities. SNP genotyping arrays have several fundamental limitations:

1. Rare Variants

SNPs must be specifically placed on the array during chip design. This means only known, common variants (typically with MAF > 1%) are included. If you carry a rare variant - one found in fewer than 1 in 100 people, or unique to your family - it will almost certainly not be on the chip and will not be detected. Since many disease-causing mutations are rare, genotyping arrays will miss them. Approximately 100 million rare variants (MAF < 1%) exist in the human population that are not captured by standard genotyping arrays.

2. Structural Variants

Large-scale changes in DNA structure - including insertions (extra DNA added), deletions (DNA removed), duplications (DNA copied), inversions (DNA flipped), and translocations (DNA moved between chromosomes) - are poorly detected by SNP arrays. These structural variants can range from a few hundred to millions of base pairs and are responsible for many genetic conditions including some forms of autism, intellectual disability, and cancer predisposition. Copy number variants (CNVs) can be partially inferred from SNP array signal intensity data, but with much lower resolution and accuracy than sequencing-based methods.

3. De Novo Mutations

Every person carries approximately 60-80 new mutations that arose during egg or sperm cell formation and were not present in either parent. These de novo mutations are by definition not in the general population and not on any genotyping chip. De novo mutations are responsible for a significant fraction of severe genetic conditions in children, including many cases of intellectual disability, autism spectrum disorders, and congenital heart defects.

4. Epigenetic Modifications

DNA methylation, histone modifications, and other epigenetic changes that influence gene expression without altering the DNA sequence are not detected by standard SNP genotyping. Epigenetic changes are involved in cancer, aging, environmental responses, and developmental disorders. Specialized microarrays (like the Illumina EPIC array) exist for methylation profiling, but they are distinct from genotyping arrays.

5. Short Tandem Repeats (STRs)

Repetitive DNA sequences where a short motif (2-6 base pairs) is repeated multiple times are not accurately measured by SNP arrays. STRs are used in forensic identification (they are the basis of DNA fingerprinting), paternity testing, and are involved in diseases like Huntington's disease (CAG repeat expansion) and Fragile X syndrome (CGG repeat expansion).

6. Mitochondrial Heteroplasmy

While SNP arrays can detect the primary mitochondrial DNA genotype for haplogroup assignment, they cannot reliably detect heteroplasmy - the condition where a person carries a mixture of different mitochondrial DNA sequences within their cells. Heteroplasmy can be clinically significant in mitochondrial diseases and is better detected by deep sequencing of mitochondrial DNA.

Quality Metrics: How Good Is the Data?

The accuracy and reliability of SNP genotyping data are measured by several key quality metrics that laboratories monitor closely:

Call Rate

The call rate is the percentage of SNPs on the array for which the algorithm could confidently assign a genotype. A high-quality sample on a modern array typically achieves a call rate of 99.5% or higher. This means that out of 700,000 SNPs, fewer than 3,500 fail to produce a confident genotype call. Samples with call rates below 98% are typically flagged for re-processing or rejected. Low call rates can result from degraded DNA, insufficient DNA quantity, or laboratory processing issues.

Concordance Rate

The concordance rate measures how often the genotyping array produces the same result when the same sample is tested multiple times, or when compared to an alternative technology (like sequencing). Modern Illumina arrays achieve concordance rates exceeding 99.9% for common SNPs. This means that if you were genotyped twice, 699,300 out of 700,000 SNPs would give identical results.

Reproducibility

Closely related to concordance, reproducibility refers to the consistency of results across different runs, different laboratories, and different chip lots. The Illumina GSA demonstrates reproducibility exceeding 99.8% in multi-site studies, making it one of the most reliable molecular assay platforms available.

GenTrain Score

Illumina's proprietary GenTrain score rates the quality of each SNP's clustering pattern on a scale of 0 to 1. SNPs with GenTrain scores above 0.7 are considered high-quality and are included in standard analyses. SNPs with scores below 0.5 are typically excluded. This metric ensures that only SNPs with clean, well-separated genotype clusters are used in downstream analysis.

The Data Processing Pipeline

After the microarray scanning is complete, the raw fluorescence intensity data goes through a sophisticated computational pipeline before it becomes the ancestry reports and health insights you see in your Helixline account:

1. Image processing: The raw scanner images are processed to identify each bead location and measure fluorescence intensity at each wavelength
2. Normalization: Intensity values are normalized to correct for technical variation across different regions of the chip, different samples on the same chip, and different chip lots
3. Genotype calling: Algorithms assign AA, AB, or BB genotypes based on the ratio and intensity of fluorescent signals, using reference cluster positions derived from thousands of previously genotyped samples
4. Quality filtering: SNPs and samples that fail quality thresholds (call rate, GenTrain score, etc.) are flagged or removed
5. Imputation (optional): Statistical algorithms can infer (impute) genotypes at millions of additional SNP positions that were not directly tested on the array, using reference panels like the 1000 Genomes Project or the Haplotype Reference Consortium. This effectively expands the dataset from ~700,000 directly genotyped SNPs to 5-10 million imputed SNPs
6. Phasing: Algorithms determine which alleles are on the same chromosome (haplotype), which is essential for accurate ancestry analysis and relative matching
7. Ancestry analysis: The phased genotype data is compared against reference panels from known populations using algorithms like ADMIXTURE, RFMix, or proprietary methods
8. Health and trait analysis: Specific genotypes at clinically-validated SNPs are interpreted according to published literature and clinical databases (ClinVar, PharmGKB)

Raw DNA Data: What You Actually Get

When you download your raw DNA data from Helixline or any other genotyping service, you receive a text file containing your genotype at each tested SNP position. A typical line in the raw data file looks like this:

• rsID: The unique identifier for the SNP (e.g., rs1426654 - the SLC24A5 skin pigmentation variant)
• Chromosome: Which chromosome the SNP is located on (e.g., chromosome 15)
• Position: The exact base-pair position on that chromosome (e.g., 48,426,484)
• Genotype: Your two alleles at that position (e.g., AA, AG, or GG)

This raw data file is typically 10-50 MB in size and contains 600,000-900,000 lines, one for each genotyped SNP. It can be uploaded to third-party analysis services for additional interpretation beyond what your original testing company provides. The raw data is yours and represents a permanent record of your genetic information at those tested positions.

The Future of Genotyping Technology

SNP genotyping technology continues to evolve, with several trends shaping its future:

• Higher density arrays: Next-generation arrays may include 1-2 million SNPs, capturing even more genetic variation including less common variants
• Population-specific optimization: Arrays designed specifically for underrepresented populations (including South Asian, African, and Indigenous populations) will improve accuracy for these groups, addressing a significant current limitation
• Integration with sequencing: Hybrid approaches that combine the cost-efficiency of genotyping with targeted sequencing of clinically important regions may become the new standard
• Declining sequencing costs: As whole genome sequencing costs approach $100, the cost advantage of genotyping narrows. Within the next decade, WGS may replace genotyping for some applications
• Long-read sequencing: Emerging technologies from Oxford Nanopore and Pacific Biosciences that read very long DNA fragments may eventually provide both SNP data and structural variant data in a single test
• Clinical integration: As genotyping data increasingly enters clinical workflows (pharmacogenomics, carrier screening), the technology will need to meet ever-higher standards of accuracy and reproducibility

Frequently Asked Questions

What is a SNP?

A SNP (single nucleotide polymorphism, pronounced "snip") is a variation at a single position in the DNA sequence where different individuals carry different nucleotide letters. For example, at a specific position, some people have an A while others have a G. SNPs are the most common type of genetic variation, occurring approximately once every 300 base pairs in the human genome. There are roughly 10 million common SNPs in the human population, and any individual carries about 4-5 million SNPs that differ from the reference genome. SNPs can influence traits, disease risk, drug response, and ancestry, making them the foundation of modern consumer DNA testing.

How many SNPs do humans have?

The human genome contains approximately 10 million common SNPs (positions where at least 1% of the population carries a different nucleotide). Any individual person typically carries 4-5 million SNPs where they differ from the human reference genome. Consumer DNA tests genotype about 600,000-900,000 of these, focusing on the most informative SNPs for ancestry, health, and pharmacogenomics. The NCBI's dbSNP database catalogs over 600 million known variants including rare ones. Of the common SNPs, approximately 20,000-25,000 fall within protein-coding regions of genes where they can directly affect protein function.

What is the difference between genotyping and sequencing?

SNP genotyping uses a microarray chip to check approximately 700,000 specific, pre-selected positions in the genome - it is like looking up specific words in a dictionary. Whole genome sequencing (WGS) reads every single one of the 3.2 billion base pairs in the genome - it is like reading the entire dictionary from cover to cover. Genotyping costs $50-200 and takes 2-4 weeks; WGS costs $200-1,000 and takes 4-8 weeks. Genotyping excels at detecting common known variants efficiently, while sequencing is necessary for finding rare mutations, structural variants, and novel genetic changes. For most consumer applications (ancestry, common traits, carrier screening), genotyping provides excellent value and sufficient information.

What can't SNP genotyping detect?

SNP genotyping has several important limitations. It cannot detect rare variants (mutations found in fewer than 1% of people or unique to your family), structural variants (large insertions, deletions, or duplications of DNA), de novo mutations (new mutations not inherited from either parent), epigenetic modifications (chemical changes to DNA that affect gene activity), short tandem repeats (repetitive DNA sequences used in forensics and involved in diseases like Huntington's), or mitochondrial heteroplasmy (mixtures of mitochondrial DNA variants). For these applications, whole genome sequencing or specialized targeted tests are required. If you have specific clinical concerns, consult a genetic counselor about which testing approach is most appropriate.

Conclusion

SNP genotyping is an elegant, efficient, and remarkably powerful technology that has democratized access to genetic information. By reading approximately 700,000 strategically selected positions across your genome, a DNA microarray chip the size of a postage stamp can determine your ancestral origins, identify your maternal and paternal lineages, detect carrier status for genetic conditions, predict drug responses, and connect you with relatives you never knew existed.

Understanding how this technology works - its principles, capabilities, and limitations - empowers you to interpret your DNA results with appropriate context. SNP genotyping is not a complete read of your genome, but for the vast majority of consumer genetics applications, it provides comprehensive, accurate, and actionable information at an accessible price point.

The technology behind your DNA test represents decades of advances in molecular biology, semiconductor engineering, and computational genomics. Every time you log into your Helixline account and explore your ancestry composition or health insights, you are benefiting from the combined work of thousands of scientists and engineers who turned the abstract concept of reading human DNA into an affordable, accessible reality.

Ready to experience this technology firsthand? Order your Helixline DNA kit and see what ~700,000 SNPs reveal about your unique genetic story.