A universal DNA barcode for the Tree of Life

We develop a novel barcoding method that uses exceptionally low-coverage genome skim data to create a varKode. We then use these varKodes to train neural networks for species identification. We optimize neural network training using a newly generated genomic data set from an expertly annotated clade of plants, the flowering plant family Malpighiaceae and its close relatives, and find nearly perfect identification for both species and genera in. We further illustrate the broad utility and discriminatory power of varKodes across additional clades of plants, fungi, animals, and bacteria.