Structural variants, including duplications, insertions, deletions, inversions, and translocations of large blocks of DNA sequence, have been shown to be associated with various human diseases. These variants also frequently occur as somatic alterations in cancer. Identifying and characterizing structural variants in a genome sequence is a challenging task. We propose to develop computational methods to enable comprehensive studies of structural variation in normal and diseased genomes. In Aim 1 we develop a general computational framework for classification and comparison of structural variants across multiple samples and measurement platforms using a novel geometric and probabilistic approach. In Aim 2 we design algorithms to maximize the effectiveness of emerging single-molecule sequencing technologies for detecting and assembling complex structural variants and rearranged transcripts. In Aim 3 we develop algorithms to reconstruct the organization of cancer genomes and investigate how structural variants alter genome organization during somatic evolution. Finally, in Aim 4, we study the population genetics of inversion polymorphisms in the human genome, including their effects on haplotype block structure and whether inversions under selection leave distinctive genetic signatures. We will apply these approaches to data from human, cancer, mouse, and pathogen genomes in collaboration with several biomedical researchers. Successful completion of the proposed studies will facilitate future research of the role of structural variation in human and cancer genetics.