What Getting Your DNA Analyzed Costs in 2026
Getting your DNA analyzed costs between $49 and about $5,000, and the spread is not mostly about quality of the lab. A consumer genotyping array runs $49 to $199. A consumer whole-genome sequence at 30x coverage runs roughly $300 to $1,200. A clinical single-gene test or targeted panel is billed at $250 to $1,500. A clinical exome or genome ordered through a physician carries a list price of $2,000 to $5,000 or more, though what you pay after insurance is often far less. What matters is the cost of the interpretation you get, the file formats you are allowed to keep, and whether the result answers a question you had. The sequencing price is not the number that matters.
What each price tier buys
A genotyping array at $49 to $199 measures 600,000 to 900,000 predetermined positions. It checks whether specific known variants are present. It does not read your genome. You get ancestry, a handful of well-studied risk alleles, and a raw text file of roughly 25 MB. Its columns are rsID, chromosome, position and genotype. Everything outside the array design is invisible, including nearly all rare variants, which are the variants most likely to matter for a specific individual.
Whole-exome sequencing at $400 to $1,000 reads the protein-coding portion of the genome, about 1 to 2 percent of the sequence, typically at 100x mean coverage on target. This is a reasonable buy if your question is Mendelian and coding. Its limitation is structural: capture probes leave uneven coverage, and regulatory variants fall outside the target entirely. Large-scale annotation work has shown that the majority of functional elements in the genome sit outside coding exons.1 The ENCODE pilot established this pattern across the first 1 percent of the genome that was surveyed densely.2
Whole-genome sequencing at 30x mean coverage, now $300 to $1,200 from consumer providers, reads the whole sequence with even coverage. You get single-nucleotide variants and small insertions and deletions. You also get copy-number changes at reasonable resolution and mitochondrial sequence. Short reads still handle repeat expansions and large structural rearrangements poorly. If your question involves a repeat disorder, short-read WGS is the wrong instrument, and you need a clinician to order the targeted assay.
Why the advertised price and the total cost differ
Sequencing is a batch process, and per-sample economics depend heavily on how many libraries share a flow cell and how much coverage each one needs. Work on cost-efficient genomic surveillance in hospital settings has quantified how multiplexing and coverage targets drive per-sample cost down by large factors without losing the analytical resolution the application needs.3 The consumer price you see is a downstream reflection of that batching. A $299 genome and a $1,200 genome may come off similar instruments with similar chemistry.
What separates the tiers is everything after the FASTQ files. Ask three questions before you buy.
First, which files do you receive? A provider that gives you only a curated PDF report and a filtered VCF has sold you their interpretation. Insist on the raw reads as FASTQ or the aligned reads as CRAM, plus a gVCF rather than a sites-only VCF. The gVCF records reference confidence at positions where no variant was called, which is the only way to distinguish “this position is reference” from “this position was never covered.”
Second, which reference build? GRCh38 is the answer you want. A large amount of older annotation is still keyed to GRCh37 or hg19, and mixing coordinate systems silently produces wrong lookups. If you receive hg19 data, liftover with CrossMap or bcftools +liftover and expect to lose a small fraction of sites.
Third, what is the mean coverage and what fraction of the callable genome hits 20x? A genome advertised as 30x with 85 percent of positions above 20x behaves very differently from one at 95 percent. Run samtools coverage and mosdepth on your own CRAM and check.
The cost of processing the data yourself
If you have the raw reads, the compute bill is small compared to the sequencing. A reasonable pipeline for a 30x human genome is alignment with bwa-mem2 mem -t 32 -K 100000000 -Y, duplicate marking with samtools markdup or gatk MarkDuplicates, then variant calling with DeepVariant using --model_type=WGS.
On a 32 to 64 vCPU cloud instance, alignment takes a few hours and DeepVariant takes a few hours more. At on-demand rates of roughly $1.50 to $3.00 per hour for that instance class, end-to-end cost per genome lands in the tens of dollars. Spot instances cut that substantially if you checkpoint. We would use nf-core/sarek to orchestrate it rather than writing your own shell scripts, because the pipeline already handles the retry and resource-scaling cases you will otherwise discover the hard way.
Storage is where people get surprised. A 30x genome is roughly 100 to 120 GB as gzipped FASTQ and 80 to 100 GB as BAM. Convert to CRAM with samtools view -C -T GRCh38.fa --output-fmt-option level=8 and you land near 20 to 30 GB. At $0.023 per GB-month in S3 Standard, a CRAM costs under a dollar a month. Cold archive tiers drop that by more than an order of magnitude. Egress is the real charge: downloading 100 GB at $0.09 per GB is about $9 each time, so move the compute to the data rather than the reverse.
The common failure modes are mundane. A CRAM without the exact reference FASTA it was compressed against is unreadable, so store the reference and its .fai alongside. An annotation run with the wrong VEP cache version produces transcript identifiers that do not match your literature lookups. A sites-only VCF will make you believe a gene is clean when it was simply not covered.
What DNA analysis does not tell you
Even a perfect genome answers a narrower set of questions than most buyers expect. Your germline sequence is fixed at conception and says nothing about current gene expression, protein abundance, or metabolic state. Cancer genomics made this concrete: the systematic characterization of glioblastoma required matched tumor and normal DNA together with expression and methylation data to identify the pathways that were disrupted.4 The same integration was necessary in clear cell renal carcinoma, where a metabolic shift showed up in expression and metabolite data rather than in DNA sequence alone.5
Comparative genomics makes the interpretation problem clear from another direction. When the mouse genome was sequenced, roughly 5 percent of the human genome turned out to be under purifying selection, well more than the coding fraction.6 Much of what constrains your biology sits in sequence we can identify as functional but cannot yet read for an individual.
So the practical answer on cost is this. If you want ancestry and a conversation starter, spend $99 on an array. If you want a durable asset you can reanalyze for decades as annotation improves, buy 30x whole-genome sequencing with raw file access, and expect to spend $300 to $1,200. If you have a specific clinical question about a heritable condition, do not buy a consumer product. See a genetic counselor or physician who can order the validated test and interpret it, because consumer sequencing is not diagnostic and should not be used that way.
Questions people also ask
How much does it cost to get your DNA analyzed? From $49 for a genotyping array to about $5,000 for a clinical genome at list price. Consumer 30x whole-genome sequencing, which is the best value for most technically minded buyers, sits at $300 to $1,200.
Is whole-genome sequencing worth the extra money over a $99 array? Yes, if you intend to work with the data. The array reads under a million preselected positions and cannot be reanalyzed for anything it did not measure. A genome can be recalled against better models and reannotated as knowledge improves.
Does insurance cover DNA analysis? Insurance frequently covers clinical genetic testing when a physician documents medical necessity, and coverage decisions vary by plan and indication. It does not cover consumer sequencing you order yourself. Ask the ordering clinician for a preauthorization and a written estimate of your out-of-pocket cost.
What does it cost to analyze the raw files after you get them? Tens of dollars in cloud compute per genome for alignment and variant calling, under a dollar a month to store a CRAM, and about $9 in egress each time you download 100 GB. The tooling is free and open source.
Footnotes
-
The ENCODE Project Consortium. An integrated encyclopedia of DNA elements in the human genome. Nature, 2012. https://doi.org/10.1038/nature11247 ↩
-
The ENCODE Project Consortium. Identification and analysis of functional elements in 1% of the human genome by the ENCODE pilot project. Nature, 2007. https://doi.org/10.1038/nature05874 ↩
-
Kady D. Waggle, Marissa Pacey Griffith, Alecia B. Rokes, et al. Methods for cost-efficient, whole genome sequencing surveillance for enhanced detection of outbreaks in a hospital setting. PLOS One, 2026. https://doi.org/10.1371/journal.pone.0355550 ↩
-
The Cancer Genome Atlas Research Network. Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature, 2008. https://doi.org/10.1038/nature07385 ↩
-
The Cancer Genome Atlas Research Network. Comprehensive molecular characterization of clear cell renal cell carcinoma. Nature, 2013. https://doi.org/10.1038/nature12222 ↩
-
Mouse Genome Sequencing Consortium. Initial sequencing and comparative analysis of the mouse genome. Nature, 2002. https://doi.org/10.1038/nature01262 ↩
Pothos builds longitudinal molecular profiles of individuals: whole-genome sequencing, single-cell RNA sequencing, and plasma proteomics, delivered as a detailed report, every data file, and stored cells. Build your profile.