ChromoVis: Feature-Rich Layouts of Chromosome Conformation Graphs

Recent chromosome conformation capture (3C) experiments produce pairwise interactions between fragments of DNA that are spatially close in the nucleus of a cell. Visualizing 3C data on the scale of the whole genome allows scientists to gain insight into chromosomal packing in the nucleus of a cell and to generate hypotheses about what kinds of genomic features (e.g. gene locations, DNA accessibility, GC content) correlate with spatial proximity. We introduce the chromosome layout problem which seeks a two-dimensional layout of a chromosome conformation graph such that: (1) an entire chromosome can be visualized as a strand of DNA together with its genomic features, and (2) pairwise spatial constraints obtained from biological experiments are respected as much as possible. Our approach treats the chromosome as a self-avoiding string or polymer while attempting to satisfy the constraints. The layouts we produce accurately depict spatial proximities observed in 3C experiments while avoiding undesirable occlusions and minimizing edge crossings.