5 min read
How much DNA do humans share with other animals?
Humans and chimpanzees are 98.8% identical across aligned DNA. Comparisons with other animals and plants often count shared genes instead of identical whole genomes.
Matic Broz Computational chemist
Humans and chimpanzees are about 98.8% identical across DNA that can be directly aligned. The familiar figures for dogs and cats answer a different question: how many human genes have a related gene in the other species.
Using one current, consistent gene-level measure, 85.5% of human protein-coding genes have a dog ortholog and 81.8% have a cat ortholog. Neither figure means that the same percentage of every DNA letter is identical.
How much DNA do humans share with other animals?
Humans and chimpanzees are 98.8% identical across DNA bases that can be directly aligned. Across more distant species, the most defensible short comparison often counts orthologous genes instead of identical DNA letters.[1][2]
The table keeps the measurement beside each number. Orthologs are genes in two species that descended from the same gene in their last common ancestor.
| Comparison | Best-supported short answer | What the percentage measures |
|---|---|---|
| Neanderthal | 99.7% | Identity in the 2010 draft-genome comparison |
| Chimpanzee | 98.8%; 91.2% (18,329 genes) | Identity in directly aligned DNA; current Ensembl ortholog presence |
| Orangutan | About 97%; 89.8% (18,050 genes) | Identity in unique, gap-free aligned sequence; current Ensembl ortholog presence |
| Mouse | 87.9% (17,684 genes); about 40% | Current Ensembl ortholog presence; fraction of the human genome alignable in the 2002 study |
| Pig | 87.5% (17,594 genes) | Current Ensembl ortholog presence |
| Cow | 87.1% (17,514 genes) | Current Ensembl ortholog presence |
| Rat | 86.4% (17,382 genes) | Current Ensembl ortholog presence |
| Dog | 85.5% (17,185 genes) | Current Ensembl ortholog presence |
| Horse | 85.1% (17,119 genes) | Current Ensembl ortholog presence |
| Elephant | 83.0% (16,694 genes) | Current Ensembl ortholog presence |
| Cat | 81.8% (16,444 genes) | Current Ensembl ortholog presence |
| Dolphin | 78.8% (15,849 genes) | Current Ensembl ortholog presence |
| Banana | 17–25% | Human protein-coding genes with orthologs, depending on method |
| Plants or trees | No single percentage | A species and comparison method must be specified |
The gene counts are a ProteinIQ calculation from Ensembl release 116, accessed July 27, 2026. They count protein-coding genes on human chromosomes 1–22, X, Y, and mitochondrial DNA with at least one Ensembl ortholog of any type in the other species. The aligned-sequence figures come from the relevant genome studies; the banana range comes from a comparison of three orthology methods.[2][3][4][7][8][9][10]
The chart compares seven mammals using the same Ensembl ortholog measure.
These values also reflect the completeness of each genome assembly and its gene annotation. A missing ortholog can mean that a gene was lost or gained during evolution, but it can also mean that the corresponding sequence has not been assembled, annotated, or recognized by the current method.
How much DNA do humans share with dogs?
About 85.5% of human protein-coding genes have an Ensembl ortholog in dogs. This is a gene-sharing percentage, not whole-genome DNA identity.[2][3]
That calculation found dog orthologs for 17,185 of 20,107 human protein-coding genes. It includes one-to-one, one-to-many, and many-to-many orthologs, then counts each human gene once.
The widely repeated 84% figure is therefore a reasonable shorthand for a gene-level comparison. The 2005 dog reference paper compared aligned and orthologous sequence across human, dog, and mouse genomes, but it did not establish a universal “84% identical DNA” result.[5]
How much DNA do humans share with cats?
About 81.8% of human protein-coding genes have an Ensembl ortholog in cats. The common claim that humans and cats share 90% of their DNA comes from a narrower 2007 gene analysis, not a whole-genome identity calculation.[2][3][6]
The 2007 cat draft covered about 65% of cat euchromatin. Within the genes eligible for that study, researchers detected cat counterparts for 90.2% of annotated human protein-coding genes. The denominator and incomplete cat assembly explain why that result differs from the current Ensembl calculation.[6]
Neither 81.8% nor 90.2% says that the same share of all human and cat DNA letters matches. The figures count whether a related gene is present, not how identical every nucleotide is within those genes or across the rest of the genome.
Why do DNA-sharing percentages differ?
DNA-sharing percentages differ because researchers can measure ortholog presence, alignment coverage, or sequence identity. These measurements use different numerators and denominators, so they should not be compared as if they were the same statistic.
Human and mouse make the distinction clear. The mouse reference-genome study found reciprocal best matches for about 80% of mouse genes, but only about 40% of the human genome could be aligned confidently to mouse at the nucleotide level.[4] Gene-rich coding sequence is easier to recognize across long evolutionary distances than much of non-coding DNA.
Three measurements account for most published percentages:
- Ortholog presence asks whether a human gene has an evolutionary counterpart in another species.
- Alignment coverage asks what fraction of one genome can be lined up with the other.
- Sequence identity asks what percentage of letters match within the aligned regions.
The table's gene-sharing percentages divide the number of unique human protein-coding genes with at least one Ensembl ortholog by 20,107 protein-coding genes. The calculation queried Ensembl BioMart release 116 separately for each animal and included one-to-one, one-to-many, and many-to-many orthologs.[2][3]
Ensembl constructs gene trees from a representative protein for each gene and reconciles those trees with the species tree to infer orthologs. Future releases can change the counts when assemblies, gene annotations, or the comparison pipeline improve.[2][3]
The method matters even more for distant species. Only 17% to 25% of human protein-coding genes had a banana ortholog in one comparison of three methods.[10] The detailed human–banana comparison explains why this is not a percentage of identical whole-genome DNA.
For a small set of homologous genes or proteins, Clustal Omega, MUSCLE5, or MAFFT can build a sequence alignment. These tools compare the sequences supplied to them; they do not produce one universal species-sharing percentage.
Insertions, deletions, duplications, and unaligned sequence lower whole-genome-style estimates. The human–chimpanzee comparison shows why 98.8%, 96%, and stricter figures near 85% can all describe different analyses.
Sources▼
- Initial sequence of the chimpanzee genome and comparison with the human genome Nature · 2005. https://www.nature.com/articles/nature04072
- Compare genes across species Ensembl · July 27, 2026. https://mart.ensembl.org/info/website/tutorials/compara.html
- Homology types Ensembl · July 23, 2026. https://mart.ensembl.org/info/genome/compara/homology_types.html
- Initial sequencing and comparative analysis of the mouse genome Nature · 2002. https://www.nature.com/articles/nature01262
- Genome sequence, comparative analysis and haplotype structure of the domestic dog Nature · 2005. https://www.nature.com/articles/nature04338
- Initial sequence and comparative analysis of the cat genome Genome Research · 2007. https://genome.cshlp.org/content/17/11/1675
- Comparative and demographic analysis of orang-utan genomes Nature · 2011. https://www.nature.com/articles/nature09687
- A Draft Sequence of the Neandertal Genome Science · 2010. https://pmc.ncbi.nlm.nih.gov/articles/PMC5100745/
- Complete Neanderthal Genome Sequenced National Human Genome Research Institute · 2010. https://www.genome.gov/27539119/2010-release-complete-neanderthal-genome-sequenced
- The Banana Conjecture Dessimoz Lab, University of Lausanne · 2020. https://lab.dessimoz.org/blog/2020/12/08/human-banana-orthologs
Founder and computational chemist, ProteinIQ
Dr. Matic Broz is the founder of ProteinIQ and a computational chemist. He completed a PhD focused on protein structure, molecular dynamics, and neural networks, and writes about structural biology and scientific software.