Long-read Sequence Assembly of Great Ape Genomes

The recent development and application of single-molecule, real-time (SMRT) sequence technology has shown considerable promise in improving human genome assemblies as well as our understanding of genetic variation. Existing great ape genome assemblies contain hundreds of thousands of gaps, concealing genetic differences that separate the great ape species. Here we present accurate assemblies of the Western lowland gorilla and common chimpanzee genomes generated from SMRT sequence data and application of a string-graph de novo assembly algorithm (Falcon). The new reference genomes of gorilla and chimpanzee represent a more than 200-fold improvement in contiguity with contig N50s of 9.6 Mbp and 12.7 Mbp, respectively. Over half a million gaps are closed with respect to the previously published gorilla and chimpanzee genome assemblies generated by a mix of ABI and Illumina sequencing data. We estimate that 98.9% and 96% of the gorilla and chimpanzee euchromatin assembled into 1854 and 1537 sequence contigs, respectively, on the basis of alignment to human (GRCh38). We identified 220,921 previously unidentified structural variants among human, chimp, and gorilla genomes, including hundreds that affect genes and thousands that affect regulatory regions. The three-way comparison of human, chimp, and gorilla allows us to assign each event to the phylogeny and specifically annotate novel lineage-specific structural variations. We have identified 24,588 human-specific structural variations. These comprehensive gorilla and chimpanzee genome assemblies provide new biological insight into genetic diversity, structural variation, gene loss, and representation of repeat structures within the great ape genomes