To improve display performance of custom tracks based on very large datasets, we have added functionality to let users store data files locally on their own web-accessible server (http, https, or ftp) and then view them as custom tracks in the Genome Browser without uploading the entire files to UCSC. The data files are stored in one of two new formats: bigBed or bigWig. These are binary formats that use an R-tree internally to index genome coordinate ranges, making it possible to selectively transfer only the subset of data needed to display regions actually viewed in the Genome Browser.
