Compare · Plot
Plot mitochondrial DNA
Any two of 1,414 genome features for 2,335 species, or one against lifespan, with statistics that update as you plot.
Live statistics
Statistics appear once the data has loaded.
Strongest correlations
Rank all 1,414 mtDNA features by how strongly they track a trait. Select a row to plot it.
Loading correlations…
How to read this table
- q is the Benjamini–Hochberg false discovery rate across all features tested against this trait in this group. With 1,414 tests here, some strong-looking ρ values appear by chance; q accounts for that. Per-gene features are tested in all species and in each class, not within orders.
- PIC r correlates changes along a dated tree (phylogenetically independent contrasts), so close relatives are not counted as independent evidence. Faded values have PIC p ≥ 0.05.
- Partial r (lifespan only) holds adult body mass constant, since big animals live longer and differ in many other ways.
- Traits are log10-transformed. "All species" pools fish, birds and mammals, which differ in both mtDNA and lifespan: check a single class before drawing conclusions. A correlation is not a cause.
How to use the plot
Pick what to plot
Choose any of 1,414 mtDNA features for each axis: the length, GC content and strand skews of every region; the frequency of each codon and amino acid, the base composition at fourfold-degenerate sites and the protein hydropathy (GRAVY) of each protein-coding gene; and the repeats, non-B DNA motifs and simple repeats of the whole genome and its major arc. Type in the picker to find one quickly, e.g. ND6 GC
or CUA
. Maximum lifespan and 10 other life-history traits from AnAge are in the picker's last tab.
Lifespan, body mass and other skewed traits start on a log scale; switch either axis with its Log toggle. Narrow the plot to one class or order, and highlight species to see where they sit.
Read the statistics
Live statistics describe exactly what is plotted: the number of species, Pearson r on the plotted (log-transformed where the axis is log) values, Spearman ρ on ranks, a two-tailed p-value and an optional least-squares fit line.
Corrected statistics appear when one axis is a life-history trait. They come from the data pipeline for every trait and feature pair in all species, 7 classes and 24 orders with at least 20 species: the Benjamini–Hochberg q-value, the phylogenetically independent contrasts (PIC) correlation on a dated tree and, for lifespan, the partial correlation with body mass held constant.
Share and reuse
Every view is a link: the address bar keeps the axes, scales, group and highlights, so Copy link shares exactly what you see. Species pages link here with their species highlighted.
For summary tables (n, mean, SD, min, median, max) of every region by class, order and family, see Statistics. Download every table from the data page, and see methods and sources for how each feature and statistic is computed.
Questions to start from
- GC content against GC skew of the whole genomeClasses separate on two composition axes: birds and mammals, fishes, reptiles and amphibians each form their own cloud.
- ND6 against the other protein-coding genes in mammalsND6 is the only protein-coding gene read from the L strand; its composition departs from the rest.
- Genome length against D-loop lengthMost of the variation in genome length sits in the control region.
- GC content of ND6 against lifespan in mammalsThe classic MitoAge question: does the composition of a gene track how long a species lives?
- CUA (leucine) codon usage against lifespan in mammalsCodon usage follows base composition; see which codons shift in long-lived species.
- Where humans sit among primatesWhole-genome GC content and GC skew of every primate, with Homo sapiens highlighted.