The First Reconstruction of Denisovans

new

Admin
Administrator

Posts: 81,510

The First Reconstruction of Denisovans Jun 13, 2020 22:15:59 GMT

Quote

Post by Admin on Jun 13, 2020 22:15:59 GMT

Figure S4B:
PCA1-10 of GenomeSTRiP biallelic duplication genotypes by sample library preparation and sequencing location.

Figure 4C:
PCA1-10 of Manta+Graphtyper deletion genotypes by sample library preparation and sequencing location.

Figure S4D:
PCA1-10 of Manta+Graphtyper insertion genotypes by sample library preparation and sequencing location.

Figure S4E:
PCA1-10 of Manta+Graphtyper inversion genotypes by sample library preparation and sequencing location.

Figure S5:
Effect of UMAP hyperparameters on observed clustering – GenomeSTRiP deletions. Left: Different n_neighbors values (increasing from top to bottom: 8,16,32,64) – keeping min_dist = 0.01 Right: Different min_dist values (increasing from top to bottom : 0.1,0.01,0.001,0.0001) – keeping n_neighbors = 16.

Admin
Administrator

Posts: 81,510

The First Reconstruction of Denisovans Jun 14, 2020 8:03:28 GMT

Quote

Post by Admin on Jun 14, 2020 8:03:28 GMT

Figure S6
Fine structure in UMAP. Zoomed plots of each continental cluster with population labels from Figure 1 in main text. Based on Manta+Graphtyper deletion genotypes.

Figure S7:
Population-Specific Variation – Each point represents a variant private to a population (n > 2) with the x-axis reflecting its frequency. Colours represent regional labels and random noise is added to aid visualization. High-frequency variants discussed in the text are highlighted.

We discovered a deletion that is private and at 54% frequency in the Central African Mbuti hunter-gatherer population that deletes SIGLEC5 without removing its adjacent paired receptor SIGLEC14 (Figure 2F). Siglecs, a family of cell-surface receptors that are expressed on immune cells, detect sialylated surface proteins expressed on host cells. Most SIGLECs act as inhibitors of leukocyte activation, but SIGLEC14 is an activating member which is thought to have evolved by gene conversion from SIGLEC5 (Angata et al., 2006). This evolution has been proposed to result in a selective advantage of combating pathogens that mimic host cells by expressing sialic acids, providing an additional activation pathway (Akkaya and Barclay 2013). The deletion we identify in the Mbuti, however, seems to remove the function of the inhibitory receptor, while keeping the activating receptor intact. This finding is surprising, as paired receptors are thought to have evolved to fine-tune immune responses; and the loss of an inhibitory receptor is hypothesized to result in immune hyperactivity and autoimmune disease (Lübbers et al., 2018).

Archaic Introgression
We genotyped our calls in the high coverage Neanderthal and Denisovan archaic genomes (Meyer et al., 2012; Prufer et al., 2017; Prufer et al., 2014), and find hundreds of variants that are exclusive to Africans and archaic genomes, suggesting that they were part of the ancestral variation that was lost in the out-of-Africa bottleneck. We then searched for common, highly stratified variants that are shared with archaic genomes but are not present in Africa, possibly resulting from adaptive introgression. We identify variants across a wide range of sizes, the smallest 63bp and largest 30kb (Table 1), and note that almost all lie within or near genes, potentially having functional consequences.

We replicated the putatively Denisovan introgressed duplication at chromosome 16p12.2 exclusive to Oceanians (Sudmant et al. 2015b). We explored the frequency of this variant in our expanded dataset within each Oceanian population, and despite all the Bouganville Islanders having significant East Asian admixture, which is not found in the Papuan Highlanders, we do not find a dilution of this variant in the former population: it is present at a remarkable and similar frequency in all three Oceanian populations (∼82%). These duplications form the most extreme regional-specific variants (Figure 2D, Figure S8), and their unusual allele distribution suggests that they may have remained at high frequencies after archaic introgression due to positive selection. We characterized this variant in more detail using fluorescent in situ hybridization (Figure 2E, Figure S12-13), and find that it consists of a region of the reference sequence that has duplicated and inserted into a gene-rich region ∼7Mb away in chr16p11.2, confirming a recent study (Hsieh et al., 2019). The selective pressure acting on this duplication and its target remain unknown and require further study; however, its similar frequency across the Oceanian populations examined contrasts with the differing frequency of the malaria-associated HBA2 deletion across Oceania, suggesting that malaria infection is unlikely to be driving the signal we see at the 16p12.2 duplication.

Admin
Administrator

Posts: 81,510

The First Reconstruction of Denisovans Jun 14, 2020 21:00:08 GMT

Quote

Post by Admin on Jun 14, 2020 21:00:08 GMT

Figure S8:
Regional-Specific Variation – Each point represents a variant private to a regional group (n > 2) with the y-axis illustrating its frequency. Random noise is added to aid visualization. The distribution reflects the ancestral diversity in Africa, the connectivity of Eurasia, the isolation & drift of the Americas and Oceania, and the separate Denisovan introgression event in Oceania. Oceania is notable for having private high-frequency variants that are all shared with the Denisovan genome and are within (bold) or near the illustrated genes, four of which are newly identified in this study (AQR, CEACAM, JAK1, ZNFR1). The Americas contain high frequency variants which are not shared with any archaic genomes, suggesting they arose and increased to high-frequency after they split from other populations. EA: East Asia, CSA: Central & South Asia, ME: Middle East.

Figure S9:
Continental (red) or Population (green) specific variants (n > 2) in the HGDP not found in 1000G or SGDP SV callsets binned by allele frequency. The same variant can be present in both distributions.

Figure S10:
Maximum allele frequency difference as a function of population differentiation. Blue line is loess fits. Left: including all populations – Right: after excluding populations with 10 samples or less. Deletions (Top), Insertions (Centre), Duplications (Right).

Figure S11:
Additional copy number expansions. Red bar illustrates region expanded. Top: Expansions in beta-Defensin genes. Centre: Expansions downstream of ARRDC5 prominent in Americans. Bottom: Expansion downstream TNFRSF1B private to Biaka.

Figure S12:
fibre-FISH of chr16 Oceanian-specific expansion shared with Denisovan genome at ∼82% frequency in all three Oceanian populations. Top: Cartoon illustration of location of original (16p12.2) and inserted site 7Mb away (16p11.2). Centre: Insertion (red) into region 7Mb away in clone RP11-368N21 (green). Bottom: Fiber-FISH of heterozygous duplication observed in metaphase (cell-line GM10540), yellow arrows show reference and red arrow shows duplication.

Figure S13:
chr16p12 Papuan-specific expansion shared with Denisovan genome in more detail. Top: Fiber-FISH illustrating the original site (top), the (inverted) insertion sites (centre) and the region surrounding the insertion site (bottom). Region flanking the insertion site (C9) is a sequence 1Mb away from the original site, consistent with GenomeSTRiP calling a second duplication at this site in perfect LD with the initial duplication. Manta also identifies a Papuan-specific inversion at this locus. This suggests a complex event involving a duplication-inverted-insertion, an inversion and a deletion. Bottom: 10X-linked reads barcode overlap in region. Longranger also identifies a complex event at this locus. Top plot shows the original site barcode overlap and the regions of structural rearrangements, including the region of C9 (on the left). Bottom shows the insertion site. Note that this region is gene rich.

We discover multiple additional high-frequency Oceanian-private variants that are shared with the Denisovan genome (Figure 2D), illustrating the separate introgression event in Oceanians and their subsequent isolation (Browning et al., 2018). A deletion within AQR, an RNA helicase gene, is present at 63% frequency and shared only with the Altai Denisovan (Figure S15). The highest expression of this gene is in EBV-transformed lymphocytes (GTEx Consortium, 2013). RNA helicases play an important role in the detection of viral RNAs and mediating the antiviral immune response, in addition to being necessary host factors for viral replication (Ranji & Boris-Lawrie, 2010). AQR has been reported to be involved in the recognition and silencing of transposable elements (Akay et al., 2017), and is known to regulate HIV-1 DNA integration (Konig et al., 2008). Two other notable Denisovan-shared deletions of high frequency are in JAK1, encoding a kinase important in cytokine signalling (44%) and CEACAM1 (also known as CD66a) a glycoprotein part of the immunoglobulin superfamily (54%).

Admin
Administrator

Posts: 81,510

The First Reconstruction of Denisovans Jun 15, 2020 6:56:37 GMT

Quote

Post by Admin on Jun 15, 2020 6:56:37 GMT

Figure S14:
Two distinct deletions present in an Mbuti sample (HGDP00450). Top: IGV screenshot of depth in the region, deletions are present but appear complex. Bottom: Loupe screenshot of the region (as in Figure 1) showing the 2 haplotypes resolved using 10x linked-reads, each carrying a different deletion. One is the Mbuti-specific variant that deletes SIGLEC5, while the other is a common global deletion that removes SIGLEC14 creating a fused gene. Lines connecting reads illustrate that they are linked, i.e. they are from the same input DNA molecule.

Figure S15:
Top: IGV screenshot of a deletion in AQR which is present at 63% frequency in Oceanian populations. The deletion is shared with the Denisovan genome but not the Neanderthals. Also shown is HGDP00551 in the bottom IGV track (homozygous for the deletion). Bottom: Loupe screenshot of the region in HGDP00542 showing the 2 haplotypes resolved using 10x linked-reads, each carrying the deletion.

In the Americas we identify a deletion, shared only with Neanderthals, that reaches ∼26% frequency in both the Surui and Pima. This variant removes an exon in MS4A1 (Figure S16), a gene encoding the B-cell differentiation antigen CD20, which plays a key role in T cell–independent antibody responses and is the target of multiple recently developed monoclonal antibodies for B-cell associated leukemias, lymphomas and autoimmune diseases (Kuijpers et al., 2010; Marshall et al., 2017). This deletion raises the possibility that therapies developed in one ethnic background might not be effective in others, and that access to individual genome sequences could guide therapy choice.

Figure S16:
Top: IGV screenshot of a deletion in an exon of MS4A1, which encodes the B-cell differentiation antigen CD20. The deletion is shared by both Neanderthals (Altai top, Vindija middle track) and American populations (reaches ∼26% in Surui and Pima). The deletion is not present in the Denisovan genome (bottom track). Bottom: Loupe screenshot of the region in HGDP01043 showing the two haplotypes resolved using 10x linked-reads, with one carrying the deletion.

Both Neanderthals and Denisovans thus appear to have contributed potentially functional structural variants to different modern human populations. As many of the identified variants are involved in immune processes (Table 1), it is tempting to speculate that they are associated with adaptation to pathogens after modern humans expanded into new environments outside of Africa.

Multiallelic Variants and Runaway Duplications
We found a dynamic range of expansion in copy numbers, with variants previously found to be biallelic containing additional copies in our more diverse dataset. Among these multiallelic copy number variants, we find intriguing examples of ‘runaway duplications’ (Handsaker et al., 2015), variants that are mostly at low copy numbers globally, but have expanded to high copy numbers in certain populations, possibly in response to regionally-restricted selection events (Figure 3).

Admin
Administrator

Posts: 81,510

The First Reconstruction of Denisovans Jun 15, 2020 22:07:43 GMT

Quote

Post by Admin on Jun 15, 2020 22:07:43 GMT

Figure 3:
Copy Number Expansions and Runaway Duplications. Red bar illustrates the location of the expansion. Additional examples are shown in Figure S11. A: Expansion in HPR in Africans and Middle Eastern samples. B: Expansions upstream OR7D2 that are mostly restricted to East Asia. The observed expansions in Central & South Asian samples are all in Hazara samples, an admixed population carrying East Asian ancestry. C: Expansions within HCAR2 which are particularly common in the Kalash population. D: Expansions in SULT1A1 which are pronounced in Oceanians (median copy number, 4; all other non-African continental groups, 2; Africa, 3). E: Expansions in ORM1/ORM2. This expansion was reported previously in Europeans (Handsaker et al., 2015); however, we find it in all regional groups and particularly in Middle Eastern populations. F: Expansions in PRB4 which are restricted to Africa and Central & South Asian samples with significant African admixture (Makrani and Sindhi).

We discover multiple expansions that are mostly restricted to African populations. The hunter-gatherer Biaka are notable for a private expansion downstream of TNFRSF1B that reaches up to 9 copies (Figure S11). We replicated the previously identified HPR expansions (Figure 3A), and find that they are present in almost all African populations in our study (Handsaker et al., 2015, Sudmant et al., 2015b). HPR encodes a haptoglobin-related protein associated with defense against trypanosome infections (Smith et al., 1995). We observe populations with the highest copy numbers to be Central and West African, consistent with the geographic distribution of the infection (Franco et al., 2014). In contrast to previous studies, we also find the expansion at lower frequencies in all Middle Eastern populations, which we hypothesize is due to recent gene flow from African populations.

We identified a remarkable expansion upstream of the olfactory receptor OR7D2 that is almost restricted to East Asia (Figure 3B), where it reaches up to 18 copies. Haplotype phasing demonstrates that many individuals contain the expansion on just one chromosome, illustrating that these alleles have mutated repeatedly on the same haplotype background. However, we identify a Han Chinese sample that has a particularly high copy number. This individual has nine copies on each chromosome, suggesting that the same expanded runaway haplotype is present twice in a single individual. This could potentially lead to an even further increase in copy number due to non-allelic homologous recombination (Handsaker et al., 2015).

We discovered expansions in HCAR2 (encoding HCA2) in Asians which are especially prominent in the Kalash group (Figure 3C), with almost a third of the population displaying an increase in copy number. HCA2 is a receptor highly expressed on adipocytes and immune cells, and has been proposed as a potential therapeutic target due to its key role in mediating anti-inflammatory effects in multiple tissues and diseases (Offermanns 2017). Another clinically-relevant expansion is in SULT1A1 (Figure 3D), which encodes a sulfotransferase involved in the metabolism of drugs and hormones (Hebbring et al., 2008). Although the copy number is polymorphic in all continental groups, the expansion is more pronounced in Oceanians.

De novo assemblies and sequences missing from the reference
We sequenced 25 samples from 13 populations using linked-read sequencing at an average depth of ∼50x and generated de novo assemblies using the Supernova assembler (Weisenfeld et al., 2017) (Table S2). By comparing our assemblies to the GRCh38 reference, we identified 1631 breakpoint-resolved unique, non-repetitive insertions across all chromosomes which in aggregate account for 1.9Mb of sequences missing from the reference (Figure 4A). A San individual contained the largest number of insertions, consistent with their high divergence from other populations. However, we note that the number of identified insertions is correlated with the assembly size and quality (Figure S18), suggesting there are still additional insertions to be discovered.

Figure S17:
Top: IGV screenshot of a small deletion (63 bp) in ZNRF1 which is present at 34% frequency in Oceanian populations. Top track Altai Neanderthal, middle track Altai Denisova, bottom track Vindija Neanderthal. The deletion is present in all 3 archaic genomes. Bottom: Loupe screenshot of the region in HGDP00542 showing the two haplotypes resolved using 10x linked-reads, with one carrying the deletion.

Figure S18:
Correlation between Contig N50 and Number of identified NUIs (r = 0.91). Colours refer to the regional group of the samples.

Figure 4:
Non-Reference Unique Insertions (NUIs). A: Ideogram illustrating the density of identified NUI locations across different chromosomes using a window size of 1 Mb. Colours on chromosomes reflect chromosomal bands with red for centromeres. B: Size distribution of NUIs using a bin size of 500bp. C: PCA of NUI genotypes showing population structure (PC3-4). Previous PCs potentially reflect variation in size and quality of the assemblies.

We find that the majority of insertions are relatively small, with a median length of 513bp (Figure 4B). They are of potential functional consequence as 10 appear to reside in exons. These genes are involved in diverse cellular processes, including immunity (NCF4), regulation of glucose (FGF21), and a potential tumour suppressor (MCC). Although many insertions are rare - 41% are found in only one or two individuals - we observe that 290 are present in over half of the samples, suggesting the reference genome may harbour rare deletion alleles at these sites. These variants show population structure, with Central Africans and Oceanians showing most differentiation (Figure 4C), reflecting the deep divergences within Africa and the effect of drift, isolation and possibly Denisovan introgression in Oceania. While the number of de novo assembled genomes using linked or long reads is increasing, they are mostly representative of urban populations. Here, we present a resource containing a diverse set of assemblies with no access or analysis restrictions.

Discussion
In this study we present a comprehensive catalogue of structural variants from a diverse set of human populations. Our analysis illustrates that a substantial amount of variation, some of which reaches high frequency in certain populations, has not been documented in previous sequencing projects. The relatively large number of high-coverage genomes in each population allowed us to identify and estimate the frequency of population-specific variants, providing insights into potentially geographically-localized selection events, although further functional work is needed to elucidate their effect. Our finding of common clinically-relevant regionally private variants, some of which appears to be introgressed from archaic hominins, argues for further efforts generating genome sequences without data restrictions from under-represented populations. We note that despite the diversity found in the HGDP panel, considerable geographic gaps remain in Africa, the Americas and Australasia.

The use of short reads in this study restricts the discovery of complex structural variants, demonstrated by recent reports which uncovered a substantially higher number of variants per individual using long-read or multi-platform technologies (Audano et al., 2019; Chaisson et al., 2019). Additionally, comparison with a mostly linear human reference formed from a composite of a few individuals, and mainly from just one person, limits accurately representing the diversity and analysis of human structural variation (Schneider et al., 2017). The identification of considerable amounts of sequences missing from the reference, in this study and others (Wong et al., 2018; Sherman et al., 2019), argues for the creation of a graph-based pan-genome that can integrate structural variation (Garrison et al., 2018). Such computational methods and further developments in long-range technologies will allow the full spectrum of human structural variation to be investigated.

Last Edit: Jun 15, 2020 22:08:30 GMT by Admin