SUMMARY In order to provide a comprehensive resource for human structural variants (SVs), we generated long-read sequence data and analyzed SVs for fifteen human genomes. We sequence resolved 99,604 insertions, deletions, and inversions including 2,238 (1.6 Mbp) that are shared among all discovery genomes with an additional 13,053 (6.9 Mbp) present in the majority, indicating minor alleles or errors in the reference. Genotyping in 440 additional genomes confirms the most common SVs in unique euchromatin are now sequence resolved. We report a ninefold SV bias toward the last 5 Mbp of human chromosomes with nearly 55% of all VNTRs (variable number of tandem repeats) mapping to this portion of the genome. We identify SVs affecting coding and noncoding regulatory loci improving annotation and interpretation of functional variation. These data provide the framework to construct a canonical human reference and a resource for developing advanced representations capable of capturing allelic diversity.

Highlights d Brain organoids preserve gene expression networks despite elevated metabolic stress d Chimpanzee organoids enable studies of the evolution of human brain development d Primary and organoid samples reveal 261 human-specific gene expression changes d Human radial glia exhibit increased mTOR activation compared to non-human primates

Genetic studies of human evolution require high-quality contiguous ape genome assemblies that are not guided by the human reference. We coupled long-read sequence assembly, full-length cDNA sequencing with a multi-platform scaffolding approach to produce ab initio chimpanzee and orangutan genome assemblies. Comparing these with two long-read de novo human genome assemblies and a gorilla genome assembly, we characterized lineage-specific and shared great ape genetic variation ranging from single base-pair to megabase-sized variants. We identified ~17 thousand fixed human-specific structural variants identifying genic and putative regulatory changes that emerged in humans since divergence from nonhuman apes. Interestingly, these fixed human-specific structural variants are enriched near genes that are downregulated in human compared to chimpanzee cerebral organoids, particularly in cells analogous to radial glial neural progenitors.

Genetic changes causing brain size expansion in human evolution have remained elusive. Notch signaling is essential for radial glia stem cell proliferation and is a determinant of neuronal number in the mammalian cortex. We find that three paralogs of human-specific NOTCH2NL are highly expressed in radial glia. Functional analysis reveals that different alleles of NOTCH2NL have varying potencies to enhance Notch signaling by interacting directly with NOTCH receptors. Consistent with a role in Notch signaling, NOTCH2NL ectopic expression delays differentiation of neuronal progenitors, while deletion accelerates differentiation into cortical neurons. Furthermore, NOTCH2NL genes provide the breakpoints in 1q21.1 distal deletion/duplication syndrome, where duplications are associated with macrocephaly and autism and deletions with microcephaly and schizophrenia. Thus, the emergence of human-specific NOTCH2NL genes may have contributed to the rapid evolution of the larger human neocortex, accompanied by loss of genomic stability at the 1q21.1 locus and resulting recurrent neurodevelopmental disorders.

We developed a computational method based on polyploid phasing of long sequence reads to resolve collapsed regions of segmental duplications within genome assemblies. The approach, Segmental Duplication Assembler (SDA), constructs graphs where paralogous sequence variants define the nodes and long-read sequences provide attraction and repulsion edges allowing us to partition and assemble long reads corresponding to distinct paralogs. We apply it to single-molecule, real-time sequence data from three human genomes and recover 33–79 Mbp of duplications where approximately half of the loci are diverged (<99.8%) when compared to the reference genome. We show that the corresponding sequence is highly accurate (>99.9%) and that the diverged sequence corresponds to copy number variable paralogs that are absent from the human reference. Our method can be applied to other complex genomes to resolve the last gene-rich gaps, improve duplicate gene annotation, and better understand copy number variant genetic diversity at the base-pair level.

Despite the importance of duplicate genes for evolutionary adaptation, accurate gene annotation is often incomplete, incorrect, or lacking in regions of segmental duplication. We developed an approach combining long-read sequencing and hybridization capture to yield full-length transcript information and confidently distinguish between nearly identical genes/paralogs. We used biotinylated probes to enrich for full-length cDNA from duplicated regions, which were then amplified, size-fractionated, and sequenced using single-molecule, long-read sequencing technology, permitting us to distinguish between highly identical genes by virtue of multiple paralogous sequence variants. We examined 19 gene families as expressed in developing and adult human brain, selected for their high sequence identity (average >99%) and overlap with human-specific segmental duplications (SDs). We characterized the transcriptional differences between related paralogs to better understand the birth-death process of duplicate genes and particularly how the process leads to gene innovation. In 48% of the cases, we find that the expressed duplicates have changed substantially from their ancestral models due to novel sites of transcription initiation, splicing, and polyadenylation, as well as fusion transcripts that connect duplication-derived exons with neighboring genes. We detect unannotated open reading frames in genes currently annotated as pseudogenes, while relegating other duplicates to nonfunctional status. Our method significantly improves gene annotation, specifically defining full-length transcripts, isoforms, and open reading frames for new genes in highly identical SDs. The approach will be more broadly applicable to genes in structurally complex regions of other genomes where the duplication process creates novel genes important for adaptive traits.

BackgroundGene innovation by duplication is a fundamental evolutionary process but is difficult to study in humans due to the large size, high sequence identity, and mosaic nature of segmental duplication blocks. The human-specific gene hydrocephalus-inducing 2, HYDIN2, was generated by a 364 kbp duplication of 79 internal exons of the large ciliary gene HYDIN from chromosome 16q22.2 to chromosome 1q21.1. Because the HYDIN2 locus lacks the ancestral promoter and seven terminal exons of the progenitor gene, we sought to characterize transcription at this locus by coupling reverse transcription polymerase chain reaction and long-read sequencing.Results5' RACE indicates a transcription start site for HYDIN2 outside of the duplication and we observe fusion transcripts spanning both the 5' and 3' breakpoints. We observe extensive splicing diversity leading to the formation of altered open reading frames (ORFs) that appear to be under relaxed selection. We show that HYDIN2 adopted a new promoter that drives an altered pattern of expression, with highest levels in neural tissues. We estimate that the HYDIN duplication occurred ~3.2 million years ago and find that it is nearly fixed (99.9%) for diploid copy number in contemporary humans. Examination of 73 chromosome 1q21 rearrangement patients reveals that HYDIN2 is deleted or duplicated in most cases.ConclusionsTogether, these data support a model of rapid gene innovation by fusion of incomplete segmental duplications, altered tissue expression, and potential subfunctionalization or neofunctionalization of HYDIN2 early in the evolution of the Homo lineage.Electronic supplementary materialThe online version of this article (doi:10.1186/s13059-017-1163-9) contains supplementary material, which is available to authorized users.

Direct comparisons of human and non-human primate brain tissue have the potential to reveal molecular pathways underlying remarkable specializations of the human brain. However, chimpanzee tissue is largely inaccessible during neocortical neurogenesis when differences in brain size first appear. To identify human-specific features of cortical development, we leveraged recent innovations that permit generating pluripotent stem cell-derived cerebral organoids from chimpanzee. First, we systematically evaluated the fidelity of organoid models to primary human and macaque cortex, finding organoid models preserve gene regulatory networks related to cell types and developmental processes but exhibit increased metabolic stress. Second, we identified 261 genes differentially expressed in human compared to chimpanzee organoids and macaque cortex. Many of these genes overlap with human-specific segmental duplications and a subset suggest increased PI3K/AKT/mTOR activation in human outer radial glia. Together, our findings establish a platform for systematic analysis of molecular changes contributing to human brain development and evolution.