Contiguity information is important to achieving high-quality de novo assembly of mammalian genomes and the haplotype-resolved resequencing of human genomes. The methods described herein pursue cost-effective, massively parallel capture of contiguity information at different scales.