Canvas

Canvas is an algorithm that calls copy number variants from a diploid sample or a matched pair of tumor and normal samples. Most of a normal DNA sample is diploid, which means having paired chromosomes. Canvas identifies regions of the sample genome that have been lost, gained, or remained the same ploidy.

Canvas scans the genome for regions with an unexpected number of short read alignments. Regions with fewer than the expected number of alignments are classified as losses. Regions with more than the expected number of alignments are classified as gains.

Canvas is intended for low-depth cytogenetics experiments, low-depth single-cell experiments, or whole-genome sequencing experiments. Canvas is not appropriate for an experiment with the following conditions:

•

Most of the genome is not assumed to be diploid.

•

Reads are not distributed randomly across the diploid genome.

For more information, visit github.com/Illumina.canvas.

For Research Use Only. Not for use in diagnostic procedures.