Human genetic adaptation related to cellular zinc homeostasis
Open accessThe DOI matches its Crossref record; 75 of the 75 references it cites were found in a registry or an index; 33 of the 36 sources its claims cite were checked for support.
Abstract
SLC30A9 encodes a ubiquitously zinc transporter (ZnT9) and has been consistently suggested as a candidate for positive selection in humans. We show it is found in homozygosis in the Denisovan genome and displays accompanying signatures suggestive of archaic introgression. The derived allele at rs1047626 is known to be associated with greater susceptibility to several neuropsychiatric traits, and we propose that adaptation to cold may have driven this selection event, while also impacting predisposition to neuropsychiatric disorders in modern humans.
Overview
CoChat Summary
## TL;DR
The paper argues that the human SLC30A9 region shows contrasting positive selection in Africa versus outside Africa, with the derived rs1047626 (Met50Val) allele associated with a Denisovan-like haplotype in non-Africans. In cell experiments, the ZnT9-50Val variant altered mitochondrial and ER zinc homeostasis and increased mitochondrial metabolism, suggesting a functional basis for selection.
## Key findings
- In SLC30A9, the authors found 18 candidate SNPs among 354 biallelic SNPs with MAF ≥ 0.02 after filtering for high differentiation and functional relevance; these included 1 nonsynonymous SNP (rs1047626), 2 synonymous variants, 5 3’ UTR variants, and 10 intronic variants.
- For the non-African lineages, CLUES estimated positive selection on rs1047626 at about 56–77 kya with selection coefficients of s = 0.0113 in CHB and s = 0.0017 in CEU.
- In Africa, the authors analyzed rs4861157 and CLUES estimated s = -0.00425 with logLR = 16.2401, for an adaptive event within the last 1,000 generations (~28,000 years BP) in YRI.
- The putatively introgressed segment around rs1047626 was 70,614 bp long; the paper reports this length is highly unlikely under incomplete lineage sorting (p-value < 2.07 × 10−8 with mutation rate μ = 1 × 10−9).
- In cell assays, ZnT9-50Val showed no difference from ZnT9-50Met in protein expression or gross localization, but it increased mitochondrial zinc content and higher metabolic activity in MTT assays under zinc-containing conditions.
- In liver samples, there were no significant differences in zinc content by rs1047626 genotype (A vs G: p = 0.107; AA/AG/GG: p = 0.322), though the G allele showed a tendency toward higher liver zinc.
## Methods
- The study combined population-genetic scans and comparative haplotype analyses using PopHuman, 1000 Genomes, HGDP, Relate, CLUES, iSAFE, S prime, U, and Q95 to assess selection and introgression at SLC30A9.
- For functional work, the authors used HEK293 and C2C12 cells transfected with ZnT9-50Met, ZnT9-50Val, or control vectors, plus imaging and biochemical assays to measure protein expression, localization, cytosolic/mitochondrial/ER zinc, and MTT-based metabolism.
- For a human tissue correlate, they genotyped 143 liver DNA samples from healthy western European-origin individuals with measured liver zinc concentrations.
## Limitations
- The paper explicitly states that the pathway from the molecular phenotype to an organism-level adaptive phenotype remains unresolved, and that the role of regulatory variants affecting SLC30A9 expression and liver zinc content cannot be ruled out.
- The authors note that their cell assays used overexpression, and that experiments at endogenous ZnT9 levels in relevant tissues could better define the effect of Met50Val.
- The paper also says more work is needed to determine the exact timing, place, and admixture events involving archaic humans.
- Evident from the text: several introgression and selection inferences rely on computational/statistical analyses of public genomic datasets, so the archaic-origin conclusions remain suggestive rather than directly proven by the available evidence.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures
![Fig 1 Signals of positive selection at SLC30A9. (A) Signatures of adaptation in the SLC30A9 region. Chromosomal location and genes are displayed at the top. In the first selection panel, window-based FST values when comparing CEU versus YRI and CHB versus YRI are represented by cyan and pink bars, respectively. Similarly, window-based Fay and Wu’s H values and XP-EHH scores (using YRI as the reference population) in CEU and CHB are shown with the same color scheme in the middle and bottom selection panels, respectively. Overlapping population scores can result in blue shades. Horizontal cyan and pink lines represent the genomic average of each test for each population, whereas 2 standard deviations are represented by the background shade in each case as a mean to interpret the genome-wide significance of each selection signature. All selection test values were extracted from the corresponding 1000 GP populations as available at the PopHuman browser (http://pophuman.uab.cat) [26] and represented using the Gviz package [58]. (B) Allele frequency trajectories and selection coefficients inferred with CLUES for rs1047626 in CHB and CEU and for rs4861157 in YRI.](https://figures.cochat.ai/paper-figures/4a18b6d2bb2741e99c05755eb5130b1f/human-genetic-adaptation-related-to-cellular-zinc-homeostasis-figure-1.jpg)
![Fig 2 Potential signatures of adaptive introgression at SLC30A9. (A) Haplotypes along the putatively introgressed region (chr4: 41,977,828–42,048,441; GRCh38) clustered and sorted by increasing distance with the Denisova genome shown in the first row. Each column corresponds to a polymorphic position with a homozygous genotype in the Denisovan (112 in total; see details in S4 and S5 Tables), each row is a phased haplotype, colored according to the population of origin of the individuals shown in the top. Note that all sites with a maximum within-population minor allele frequency below 0.05 were removed. Black cells represent the presence of the derived allele and white cells the ancestral allele (see Materials and Methods). The grey box highlights the location of rs1047626. (B) Number of SNP differences that the Neanderthal and modern human haplotypes present to the Denisovan genome sorted by decreasing similarity as obtained by Haplostrips [38]. (C) Genomewide distribution of S prime introgression values in Europeans and East Asians when using either YRI (values obtained from [31]) or Mbuti as an unadmixed group. The red dot denoted the SLC30A9 region. (D) Joint distribution of the U (10%, 50%, 100%) and Q95 (10%, 100%) statistics in CEU and CHB when using either YRI or ESN as unadmixed group. Circles denote the U (10%, 50%, 100%) and Q95 (10%, 100%) values for the SLC30A9 region in each corresponding analysis, which are also shown in the inserted table together with the 99.9th and 99th genomewide percentiles.](https://figures.cochat.ai/paper-figures/70628c7c6a68459cb0f2a8646edbf758/human-genetic-adaptation-related-to-cellular-zinc-homeostasis-figure-2.jpg)



Verification
Claims the paper makes
16 Valid · 1 Not valid · 2 Not checked · 122 Own results and interpretations
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“iSAFE [28] similarly pinpointed several eQTLs (including rs1047626) among the most likely candidate mutations favored by selection along the region (chr4: 41,853,671–42,153,671; GRCh37) in both CEU and CHB”
§ Human genetic adaptation related to cellular zinc homeostasis > Results > Signals of positive selection outside Africa
What the sources say
- Could not verifycites [28] A Akbari, 2018, Identifying the favored mutation in a positive selective sweep., Nat Methods. sourceExperimental — accuracy not yet measured
The paper’s iSAFE evaluation discusses ranked candidate mutations at selected loci, but the passages in our copy do not identify rs1047626 or report several eQTLs as top candidates in the specified chromosome 4 region in both CEU and CHB. Searches for the variant, coordinates, and eQTL terminology found no passage addressing that specific result. (LLM, conf=0.73)
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
Every source the paper cites for this claim was checked and supports it.
In the paper
“we next computed the S prime statistic [31] in the Europeans and East Asian samples available in the Human Genome Diversity Panel (HGDP-CEPH) [37], using the Mbuti as an outgroup, and detected the SLC30A9 as a candidate for introgression in both groupings”
§ Human genetic adaptation related to cellular zinc homeostasis > Results > Potential signals of archaic introgression We next explored whether such an early selection event could be explained by adaptive introgression. For that, we first investigated whether the Met50Val substitution was present in publicly available Neanderthal and Denisovan genomes. Remarkably, the Denisovan genome presents the homozygous genotype for the derived allele at the rs1047626 position, whereas three high coverage Neanderthal genomes (Altai, Vindija, and Chagyrskaya) are homozygotes for the ancestral allele. We then analyzed the corresponding Neanderthal and Denisovan base positions for a total of 170 SNPs found in high LD (r2>0.8) with rs1047626 in CHB and CEU and spanning a 98.74 kb region (S4 Table). For 164 of these SNPs (96%), the most frequent allele in EAS, EUR, SAS, and AMR differs from the most prevalent in AFR, confirming the existence of two major contrasting Yin/Yang human haplotypes (i.e., two haplotypes differing at each polymorphic site; [30]). On the corresponding 98 covered SNP positions, the Neanderthal and Denisovan genomes were found to display higher allelic similarity and greater sharing of derived alleles with the allelic combinations most frequently found in non-Africans (S23 Fig and S4 Table). Given these results, we next investigated whether this region had been identified in genome scans of Neanderthal and Denisovan introgression into modern humans. No hit was found within the most probable candidate genomic regions for Neanderthal introgression [31–33]. However, two independent studies recently reported two overlapping candidate regions for Denisovan introgression in Melanesians [34] and Papua New Guineans [35], respectively (S23 and S24 Figs and S4 Table). The two most distant SNP positions where the archaic genomes were found to carry derived alleles shared with the most frequent allele combinations in present-day humans and coinciding with at least one of these introgression maps delimitate a region of 70.61 kb (chr4: 41,977,828–42,048,441; GRCh38). Consequently, we further explored the haplotype structure of this putatively introgressed region using all the SNP variants in the CHB, CEU and YRI populations from the 1000 GP [36] together with those of Oceanian individuals available in the HGDP dataset [37] presenting a homozygous genotype in the Denisovan genome, and including the corresponding nucleotide positions in the three high-coverage Neanderthal genomes previously used (see details in Materials and Methods). The resulting set of 112 SNPs defined 4 archaic haplotypes and a total of 61 modern human haplotypes with two major allelic configurations differing at 66 SNP positions: htI, present in 53% of the YRI, 13% in CEU and 2% in CHB chromosomes, and ht II, found in 95% of the CHB, 67% of the CEU, and in 3% of the YRI chromosomes. Notably, the two most frequent haplotypes in Oceanians (ht LX and ht LXI found in 52% and 30% of the Oceanian chromosomes, respectively) only show 2 and 3 SNP differences when compared with ht II (S5 and S6 Tables and see haplotype network in S25 Fig). We then used haplostrips [38] to better explore the haplotype structure of this region (chr4: 41,977,828–42,048,441; GRCh38) and genetic differences between modern human or Neanderthal haplotypes and the Denisovan genome. Notably, most of the Oceanian, CHB and CEU chromosomes (but also 8 out of 216 YRI chromosomes that carry the derived allele at rs1047626) displayed between 11 and 13 differences to the Denisova. In contrast, while the three Neanderthal haplotypes present between 17 to 28 differences to the Denisovan genome, the most common haplotype in YRI displays a total of 78 changes (Fig 2 and S7 Table). Moreover, the most frequent allelic configuration in CHB and CEU contains 77 derived alleles shared with the Denisovan genome, while the most prevalent in YRI only presents 11 derived alleles shared with archaic humans (Fig 2A and S5 Table). Similarly, the Neanderthal haplotypes, which all present the ancestral allele for rs1047626, share a minimum of 72 derived alleles with the Denisovan genome. To gain further insight into the genetic structure of such high-frequency Denisovan-like haplotype in modern humans and understand its presence in up to 7% of the YRI chromosomes, we repeated the same analyses including 4 additional Sub-Saharan African populations from the 1000 GP (i.e., ESN, MSL, GWD, and LWK; S26 Fig and S8 Table) and extending the explored region (chr4: 41,905,811–42,146,424; GRCh37; S27 Fig and S9 Table). Visual inspection of the corresponding haplostrips plots confirms that the Denisovan-like haplotypes carrying the derived Val allele in ESN (found in 4% of the ESN chromosomes), MSL (11%), GWD (10%) and LWK (8%) present the same archaic-like haplotype structure (and length ~70,614 bp) found in CHB and CEU (see S26 and S27 Figs). By contrast, the African haplotypes first clustered to the Denisovan and Altai Neanderthal configurations but presenting the ancestral allele at rs1047626, display fewer shared derived alleles with the Denisovan genome (besides that of the rs1047626 position) and seem to present a much-dispersed pattern of private and shared derived alleles with both archaic genomes that expands outside the SLC30A9 region and appears reshuffled in other African haplotypes (see S27 Fig and S9 Table). As for the Denisovan-like modern human haplotypes containing the Met to Val substitution, since the probability of observing one haplotype of 70,614 bp in length shared by ancestry without being broken by recombination given the local recombination rate of the region (1.58 cM/Mb) and the time since divergence between modern and archaic humans (see details in Materials and Methods) is highly unlikely (p-value < 2.07 x 10−8 when considering a mutation rate μ = 1 x 10−9; S10 Table), we can reject an scenario of incomplete lineage sorting (ILS) for causing the unusual haplotype configuration observed around the derived allele at rs1047626. The presence of the putatively introgressed haplotype in YRI may explain why previous scans of Denisovan introgression have not detected SLC30A9 as a candidate region for archaic introgression in Eurasians. Since Mbuti do not present the derived allele of the Met to Val substitution (S28 Fig) and probably have not been reached by the back to Africa migration [39,40], we next computed the S prime statistic [31] in the Europeans and East Asian samples available in the Human Genome Diversity Panel (HGDP-CEPH) [37], using the Mbuti as an outgroup, and detected the SLC30A9 as a candidate for introgression in both groupings (Fig 2C). Finally, we also investigated the scenario of archaic introgression in CHB and CEU with the Q95 and U statistics [41] using both the African YRI and ESN population as outgroups and considering the 10% outgroup cutoff to account for the known presence of this putatively introgressed haplotype in Africa (Fig 2D and S11 Table). In CHB, Q95YRI, DEN (10%, 100%) and Q95ESN, DEN (10%, 100%) detected unusual frequencies for the derived alleles that are shared with the Denisovan genome (and at <10% frequency in YRI or ESN, respectively) when considering the 99.9th percentile of the statistic computed genomewide; while in CEU, Q95YRI, DEN (10%, 100%) and Q95ESN, DEN (10%, 100%) were only significant when considering the 99th percentile (Fig 2D). Similarly, the U statistic indicated an excess of shared alleles with the Denisova at more than 50% frequency in CHB and CEU (while being found at <10% frequency in YRI or ESN, respectively) only when considering the 99th percentile (Fig 2D and S11 Table, and see joint U and Q95 distributions indicating the SLC30A9 region together with two previously described candidate regions for Denisovan introgression in CHB in S29 Fig for comparison). Thus, overall, the numbers and frequencies of shared Denisovan alleles found in CHB and CEU provide further suggestive evidence for a likely scenario of adaptive introgression in the SCL30A9 region.
What the sources say
- Supportedcites [31] SR Browning, 2018, Analysis of Human Sequence Data Reveals Two Pulses of Archaic Denisovan Admixture, Cell source
The citation [31] identifies the S-prime method, not the citing authors’ HGDP-CEPH analysis or their SLC30A9 finding. In Results (Simulation results), the source explicitly calls its method “Sprime”; the Introduction says it develops an S*-like method for detecting introgressed haplotypes. (LLM, conf=0.92)
- Supportedcites [37] A Bergström, 2020, Insights into human genetic variation and population history from 929 diverse genomes, Science source
Citation [37] identifies the HGDP-CEPH samples used by the citing authors. The source’s Materials and methods summary (page not provided) states that it sequenced DNA from the HGDP-CEPH panel. The S-prime analysis and SLC30A9 finding are presented as the citing authors’ own work, not as findings attributed to this source. (LLM, conf=0.93)
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“two independent studies recently reported two overlapping candidate regions for Denisovan introgression in Melanesians [34] and Papua New Guineans [35], respectively”
§ Human genetic adaptation related to cellular zinc homeostasis > Results > Potential signals of archaic introgression We next explored whether such an early selection event could be explained by adaptive introgression. For that, we first investigated whether the Met50Val substitution was present in publicly available Neanderthal and Denisovan genomes. Remarkably, the Denisovan genome presents the homozygous genotype for the derived allele at the rs1047626 position, whereas three high coverage Neanderthal genomes (Altai, Vindija, and Chagyrskaya) are homozygotes for the ancestral allele. We then analyzed the corresponding Neanderthal and Denisovan base positions for a total of 170 SNPs found in high LD (r2>0.8) with rs1047626 in CHB and CEU and spanning a 98.74 kb region (S4 Table). For 164 of these SNPs (96%), the most frequent allele in EAS, EUR, SAS, and AMR differs from the most prevalent in AFR, confirming the existence of two major contrasting Yin/Yang human haplotypes (i.e., two haplotypes differing at each polymorphic site; [30]). On the corresponding 98 covered SNP positions, the Neanderthal and Denisovan genomes were found to display higher allelic similarity and greater sharing of derived alleles with the allelic combinations most frequently found in non-Africans (S23 Fig and S4 Table). Given these results, we next investigated whether this region had been identified in genome scans of Neanderthal and Denisovan introgression into modern humans. No hit was found within the most probable candidate genomic regions for Neanderthal introgression [31–33]. However, two independent studies recently reported two overlapping candidate regions for Denisovan introgression in Melanesians [34] and Papua New Guineans [35], respectively (S23 and S24 Figs and S4 Table). The two most distant SNP positions where the archaic genomes were found to carry derived alleles shared with the most frequent allele combinations in present-day humans and coinciding with at least one of these introgression maps delimitate a region of 70.61 kb (chr4: 41,977,828–42,048,441; GRCh38). Consequently, we further explored the haplotype structure of this putatively introgressed region using all the SNP variants in the CHB, CEU and YRI populations from the 1000 GP [36] together with those of Oceanian individuals available in the HGDP dataset [37] presenting a homozygous genotype in the Denisovan genome, and including the corresponding nucleotide positions in the three high-coverage Neanderthal genomes previously used (see details in Materials and Methods). The resulting set of 112 SNPs defined 4 archaic haplotypes and a total of 61 modern human haplotypes with two major allelic configurations differing at 66 SNP positions: htI, present in 53% of the YRI, 13% in CEU and 2% in CHB chromosomes, and ht II, found in 95% of the CHB, 67% of the CEU, and in 3% of the YRI chromosomes. Notably, the two most frequent haplotypes in Oceanians (ht LX and ht LXI found in 52% and 30% of the Oceanian chromosomes, respectively) only show 2 and 3 SNP differences when compared with ht II (S5 and S6 Tables and see haplotype network in S25 Fig). We then used haplostrips [38] to better explore the haplotype structure of this region (chr4: 41,977,828–42,048,441; GRCh38) and genetic differences between modern human or Neanderthal haplotypes and the Denisovan genome. Notably, most of the Oceanian, CHB and CEU chromosomes (but also 8 out of 216 YRI chromosomes that carry the derived allele at rs1047626) displayed between 11 and 13 differences to the Denisova. In contrast, while the three Neanderthal haplotypes present between 17 to 28 differences to the Denisovan genome, the most common haplotype in YRI displays a total of 78 changes (Fig 2 and S7 Table). Moreover, the most frequent allelic configuration in CHB and CEU contains 77 derived alleles shared with the Denisovan genome, while the most prevalent in YRI only presents 11 derived alleles shared with archaic humans (Fig 2A and S5 Table). Similarly, the Neanderthal haplotypes, which all present the ancestral allele for rs1047626, share a minimum of 72 derived alleles with the Denisovan genome. To gain further insight into the genetic structure of such high-frequency Denisovan-like haplotype in modern humans and understand its presence in up to 7% of the YRI chromosomes, we repeated the same analyses including 4 additional Sub-Saharan African populations from the 1000 GP (i.e., ESN, MSL, GWD, and LWK; S26 Fig and S8 Table) and extending the explored region (chr4: 41,905,811–42,146,424; GRCh37; S27 Fig and S9 Table). Visual inspection of the corresponding haplostrips plots confirms that the Denisovan-like haplotypes carrying the derived Val allele in ESN (found in 4% of the ESN chromosomes), MSL (11%), GWD (10%) and LWK (8%) present the same archaic-like haplotype structure (and length ~70,614 bp) found in CHB and CEU (see S26 and S27 Figs). By contrast, the African haplotypes first clustered to the Denisovan and Altai Neanderthal configurations but presenting the ancestral allele at rs1047626, display fewer shared derived alleles with the Denisovan genome (besides that of the rs1047626 position) and seem to present a much-dispersed pattern of private and shared derived alleles with both archaic genomes that expands outside the SLC30A9 region and appears reshuffled in other African haplotypes (see S27 Fig and S9 Table). As for the Denisovan-like modern human haplotypes containing the Met to Val substitution, since the probability of observing one haplotype of 70,614 bp in length shared by ancestry without being broken by recombination given the local recombination rate of the region (1.58 cM/Mb) and the time since divergence between modern and archaic humans (see details in Materials and Methods) is highly unlikely (p-value < 2.07 x 10−8 when considering a mutation rate μ = 1 x 10−9; S10 Table), we can reject an scenario of incomplete lineage sorting (ILS) for causing the unusual haplotype configuration observed around the derived allele at rs1047626. The presence of the putatively introgressed haplotype in YRI may explain why previous scans of Denisovan introgression have not detected SLC30A9 as a candidate region for archaic introgression in Eurasians. Since Mbuti do not present the derived allele of the Met to Val substitution (S28 Fig) and probably have not been reached by the back to Africa migration [39,40], we next computed the S prime statistic [31] in the Europeans and East Asian samples available in the Human Genome Diversity Panel (HGDP-CEPH) [37], using the Mbuti as an outgroup, and detected the SLC30A9 as a candidate for introgression in both groupings (Fig 2C). Finally, we also investigated the scenario of archaic introgression in CHB and CEU with the Q95 and U statistics [41] using both the African YRI and ESN population as outgroups and considering the 10% outgroup cutoff to account for the known presence of this putatively introgressed haplotype in Africa (Fig 2D and S11 Table). In CHB, Q95YRI, DEN (10%, 100%) and Q95ESN, DEN (10%, 100%) detected unusual frequencies for the derived alleles that are shared with the Denisovan genome (and at <10% frequency in YRI or ESN, respectively) when considering the 99.9th percentile of the statistic computed genomewide; while in CEU, Q95YRI, DEN (10%, 100%) and Q95ESN, DEN (10%, 100%) were only significant when considering the 99th percentile (Fig 2D). Similarly, the U statistic indicated an excess of shared alleles with the Denisova at more than 50% frequency in CHB and CEU (while being found at <10% frequency in YRI or ESN, respectively) only when considering the 99th percentile (Fig 2D and S11 Table, and see joint U and Q95 distributions indicating the SLC30A9 region together with two previously described candidate regions for Denisovan introgression in CHB in S29 Fig for comparison). Thus, overall, the numbers and frequencies of shared Denisovan alleles found in CHB and CEU provide further suggestive evidence for a likely scenario of adaptive introgression in the SCL30A9 region.
What the sources say
- Supportedcites [34] G Gower, 2021, Detecting adaptive introgression in human evolution using convolutional neural networks, Elife source
In Results, under ‘Candidates for Denisovan adaptive introgression in Melanesian genomes’ (page unavailable), Gower et al. report a candidate region near SUMF1 and LRNN1 and state that Jacobs et al. (2019) also identified it. This supports the part attributed to Gower et al.: a Melanesian Denisovan-introgression candidate overlapping a region reported by the other study. (LLM, conf=0.88)
- Partly supportedcites [35] MJ Hubisz, 2020, Mapping gene flow between ancient hominins through demography-aware inference of the ancestral recombination graph, PLoS Genet. sourceExperimental — accuracy not yet measured
The Results section ('ARGweaver-D can accurately identify archaic introgression in modern humans'; page unavailable) reports predicted introgressed regions in a Papuan individual, including Denisovan sequence. It does not identify two specific candidate regions overlapping those reported by a separate study of Melanesians, so that more specific attribution is unsupported. (LLM, conf=0.82)
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“No hit was found within the most probable candidate genomic regions for Neanderthal introgression [31–33].”
§ Human genetic adaptation related to cellular zinc homeostasis > Results > Potential signals of archaic introgression We next explored whether such an early selection event could be explained by adaptive introgression. For that, we first investigated whether the Met50Val substitution was present in publicly available Neanderthal and Denisovan genomes. Remarkably, the Denisovan genome presents the homozygous genotype for the derived allele at the rs1047626 position, whereas three high coverage Neanderthal genomes (Altai, Vindija, and Chagyrskaya) are homozygotes for the ancestral allele. We then analyzed the corresponding Neanderthal and Denisovan base positions for a total of 170 SNPs found in high LD (r2>0.8) with rs1047626 in CHB and CEU and spanning a 98.74 kb region (S4 Table). For 164 of these SNPs (96%), the most frequent allele in EAS, EUR, SAS, and AMR differs from the most prevalent in AFR, confirming the existence of two major contrasting Yin/Yang human haplotypes (i.e., two haplotypes differing at each polymorphic site; [30]). On the corresponding 98 covered SNP positions, the Neanderthal and Denisovan genomes were found to display higher allelic similarity and greater sharing of derived alleles with the allelic combinations most frequently found in non-Africans (S23 Fig and S4 Table). Given these results, we next investigated whether this region had been identified in genome scans of Neanderthal and Denisovan introgression into modern humans. No hit was found within the most probable candidate genomic regions for Neanderthal introgression [31–33]. However, two independent studies recently reported two overlapping candidate regions for Denisovan introgression in Melanesians [34] and Papua New Guineans [35], respectively (S23 and S24 Figs and S4 Table). The two most distant SNP positions where the archaic genomes were found to carry derived alleles shared with the most frequent allele combinations in present-day humans and coinciding with at least one of these introgression maps delimitate a region of 70.61 kb (chr4: 41,977,828–42,048,441; GRCh38). Consequently, we further explored the haplotype structure of this putatively introgressed region using all the SNP variants in the CHB, CEU and YRI populations from the 1000 GP [36] together with those of Oceanian individuals available in the HGDP dataset [37] presenting a homozygous genotype in the Denisovan genome, and including the corresponding nucleotide positions in the three high-coverage Neanderthal genomes previously used (see details in Materials and Methods). The resulting set of 112 SNPs defined 4 archaic haplotypes and a total of 61 modern human haplotypes with two major allelic configurations differing at 66 SNP positions: htI, present in 53% of the YRI, 13% in CEU and 2% in CHB chromosomes, and ht II, found in 95% of the CHB, 67% of the CEU, and in 3% of the YRI chromosomes. Notably, the two most frequent haplotypes in Oceanians (ht LX and ht LXI found in 52% and 30% of the Oceanian chromosomes, respectively) only show 2 and 3 SNP differences when compared with ht II (S5 and S6 Tables and see haplotype network in S25 Fig). We then used haplostrips [38] to better explore the haplotype structure of this region (chr4: 41,977,828–42,048,441; GRCh38) and genetic differences between modern human or Neanderthal haplotypes and the Denisovan genome. Notably, most of the Oceanian, CHB and CEU chromosomes (but also 8 out of 216 YRI chromosomes that carry the derived allele at rs1047626) displayed between 11 and 13 differences to the Denisova. In contrast, while the three Neanderthal haplotypes present between 17 to 28 differences to the Denisovan genome, the most common haplotype in YRI displays a total of 78 changes (Fig 2 and S7 Table). Moreover, the most frequent allelic configuration in CHB and CEU contains 77 derived alleles shared with the Denisovan genome, while the most prevalent in YRI only presents 11 derived alleles shared with archaic humans (Fig 2A and S5 Table). Similarly, the Neanderthal haplotypes, which all present the ancestral allele for rs1047626, share a minimum of 72 derived alleles with the Denisovan genome. To gain further insight into the genetic structure of such high-frequency Denisovan-like haplotype in modern humans and understand its presence in up to 7% of the YRI chromosomes, we repeated the same analyses including 4 additional Sub-Saharan African populations from the 1000 GP (i.e., ESN, MSL, GWD, and LWK; S26 Fig and S8 Table) and extending the explored region (chr4: 41,905,811–42,146,424; GRCh37; S27 Fig and S9 Table). Visual inspection of the corresponding haplostrips plots confirms that the Denisovan-like haplotypes carrying the derived Val allele in ESN (found in 4% of the ESN chromosomes), MSL (11%), GWD (10%) and LWK (8%) present the same archaic-like haplotype structure (and length ~70,614 bp) found in CHB and CEU (see S26 and S27 Figs). By contrast, the African haplotypes first clustered to the Denisovan and Altai Neanderthal configurations but presenting the ancestral allele at rs1047626, display fewer shared derived alleles with the Denisovan genome (besides that of the rs1047626 position) and seem to present a much-dispersed pattern of private and shared derived alleles with both archaic genomes that expands outside the SLC30A9 region and appears reshuffled in other African haplotypes (see S27 Fig and S9 Table). As for the Denisovan-like modern human haplotypes containing the Met to Val substitution, since the probability of observing one haplotype of 70,614 bp in length shared by ancestry without being broken by recombination given the local recombination rate of the region (1.58 cM/Mb) and the time since divergence between modern and archaic humans (see details in Materials and Methods) is highly unlikely (p-value < 2.07 x 10−8 when considering a mutation rate μ = 1 x 10−9; S10 Table), we can reject an scenario of incomplete lineage sorting (ILS) for causing the unusual haplotype configuration observed around the derived allele at rs1047626. The presence of the putatively introgressed haplotype in YRI may explain why previous scans of Denisovan introgression have not detected SLC30A9 as a candidate region for archaic introgression in Eurasians. Since Mbuti do not present the derived allele of the Met to Val substitution (S28 Fig) and probably have not been reached by the back to Africa migration [39,40], we next computed the S prime statistic [31] in the Europeans and East Asian samples available in the Human Genome Diversity Panel (HGDP-CEPH) [37], using the Mbuti as an outgroup, and detected the SLC30A9 as a candidate for introgression in both groupings (Fig 2C). Finally, we also investigated the scenario of archaic introgression in CHB and CEU with the Q95 and U statistics [41] using both the African YRI and ESN population as outgroups and considering the 10% outgroup cutoff to account for the known presence of this putatively introgressed haplotype in Africa (Fig 2D and S11 Table). In CHB, Q95YRI, DEN (10%, 100%) and Q95ESN, DEN (10%, 100%) detected unusual frequencies for the derived alleles that are shared with the Denisovan genome (and at <10% frequency in YRI or ESN, respectively) when considering the 99.9th percentile of the statistic computed genomewide; while in CEU, Q95YRI, DEN (10%, 100%) and Q95ESN, DEN (10%, 100%) were only significant when considering the 99th percentile (Fig 2D). Similarly, the U statistic indicated an excess of shared alleles with the Denisova at more than 50% frequency in CHB and CEU (while being found at <10% frequency in YRI or ESN, respectively) only when considering the 99th percentile (Fig 2D and S11 Table, and see joint U and Q95 distributions indicating the SLC30A9 region together with two previously described candidate regions for Denisovan introgression in CHB in S29 Fig for comparison). Thus, overall, the numbers and frequencies of shared Denisovan alleles found in CHB and CEU provide further suggestive evidence for a likely scenario of adaptive introgression in the SCL30A9 region.
What the sources say
- Supportedcites [31] SR Browning, 2018, Analysis of Human Sequence Data Reveals Two Pulses of Archaic Denisovan Admixture, Cell source
The Results section (“Signals of positive selection”; page not provided) identifies high-frequency genomic regions matching Neanderthal or Denisovan genomes and states that all the selected regions appear to be Neanderthal-introgressed. The citation supports the identification of candidate regions; “no hit was found” describes the citing authors’ own search, not a finding attributed to this paper. (LLM, conf=0.88)
- Partly supportedcites [32] S Sankararaman, 2016, The Combined Landscape of Denisovan and Neanderthal Ancestry in Present-Day Humans, Curr Biol sourceExperimental — accuracy not yet measured
Our copy of Sankararaman et al. (2016) does supply the thing the citation is attached to: the Results section 'Regions with elevated proportions of archaic ancestry' reports scanning the introgression maps and identifying '238 windows with elevated Neanderthal ancestry in a pool of all non-Africans', framed as candidate loci of archaic positive selection. But the claim's actual assertion — that no hit was found in those candidate regions — is a null result from the citing study's own analysis; no passage in our copy (searched: 'no hit', candidate/adaptive introgression regions, association tests) reports such an absence of signal. The paper instead reports positive enrichments (keratin-filament genes enriched for Neanderthal ancestry; FOXP2 as a Neanderthal/Denisovan ancestry desert) and says formal testing of selection at elevated-ancestry regions is challenging. (LLM, conf=0.68)
- Partly supportedcites [33] L Skov, 2020, The nature of Neanderthal introgression revealed by 27,566 Icelandic genomes., Nature sourceExperimental — accuracy not yet measured
In “A diverse set of introgressed fragments” (page unavailable), the paper identifies genomic fragments with a greater than 90% posterior probability of archaic origin, supporting the reference to candidate introgression regions. It does not report that no hit was found within those regions; that part of the citing sentence is unsupported by this source. (LLM, conf=0.78)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“The motivation for this study was the previous detection of strong signatures of positive selection in the SLC30A9 region [13,14,16,17,20,25].”
§ Human genetic adaptation related to cellular zinc homeostasis > Discussion The motivation for this study was the previous detection of strong signatures of positive selection in the SLC30A9 region [13,14,16,17,20,25]. In this context, we first re-examined the genetic evidence for a hard-selective sweep using the FST, XP-EHH, and Fay and Wu’s H selection statistics and found several functional variants on SLC30A9 that could explain the strong adaptive signals detected. In accordance with Zhang et al. (2015) [13], we confirmed contrasting signals of positive selection between two major haplotypes in Africa and East Asia, the latter also being present at intermediate-high frequencies in South Asia, Europe, and America. The major African haplotype carries alleles at several eQTLs associated with higher SLC30A9 expression. In contrast, the most prevalent allelic combinations outside Africa are associated with lower gene expression and are linked to the derived allele of the rs1047626 polymorphism, which causes a Met to Val substitution at ZnT9 codon position 50. Although no experimental characterization of its corresponding molecular phenotype has been reported to date, previous studies have attributed the strong signals of adaptation found in East Asians to this non-synonymous variant [13,17,23]. In this work, we describe the presence of the derived Val allele in the Denisovan genome and an unusually high sharing of derived alleles between the major haplotype outside Africa and archaic humans (both Neanderthals and Denisovans). Moreover, we demonstrate that the persistence of such similar allelic configuration around the Met50Val substitution comprising a segment of 70,614 bp in length between modern and archaic humans cannot be explained by ILS when considering the recombination rate of the SLC30A9 region. These results are compatible with two independent studies reporting a Denisovan adaptive introgressed region in modern Melanesians [34] and Papua New Guineans [35] overlapping with the SLC30A9 gene and comprising the rs1047626 polymorphism. Interestingly, the putatively introgressed Oceanian chromosomes not only present the derived allele at rs1047626 but display haplotypes just 2–3 mutational steps away from those positively selected in CHB and CEU. However, some African chromosomes (4% in ESN, 7% in YRI, 8% in LWK, 10% in GWD, and 11% in MSL) carry the derived rs1047626 allele and present similar allelic configurations to those suggested as putatively introgressed. The presence of such Denisovan-like haplotypes in Africa could result from back-to-Africa migrations [44] and, as shown by our S prime analysis when using Mbuti as unadmixed group, it might have easily prevented the detection of the SLC30A9 gene in previous maps of archaic introgression in modern humans. Similarly, the use of the U and Q95 introgression statistics with the 10% unadmixed outgroup frequency cut-off confirmed that the SLC30A9 region displays unusual sharing of derived alleles with the Denisova as well as unusual numbers of shared Denisovan alleles at frequencies >50% in both CHB (above the 99.9th genomewide percentile for Q95) and CEU (when considering the 99th percentile), suggesting again a likely scenario for adaptive introgression. However, we also describe high similarity (but greater SNP differences) between the Denisovan genome and (at least two) Neanderthal haplotypes and other much less frequent modern human haplotypes carrying the ancestral allele at rs1047626. But while the Denisovan-like haplotypes carrying the derived allele at rs1047626 present very similar specific allelic configurations and lengths across populations, these other archaic-like configurations are more diverse and appear reshuffled in different chromosomal backgrounds, especially in Africa. In the future, the availability of new high-coverage archaic genomes and the consideration of multiple evolutionary event models may help to better explain such unusual sharing of archaic-like haplotypes in the SLC30A9 region and test the likelihood of adaptive introgression from either Denisovans, Neanderthals, or other archaic humans. For example, although we have rejected the possibility of ILS, we cannot discard a scenario of shared ancestral variation coupled with two contrasting selecting sweeps in and outside Africa creating such an unusual putative Denisovan-like configuration outside Africa. Similarly, we cannot discard other more complex scenarios including a first introgression event from the ancestors of modern humans into archaic humans followed by a second archaic introgression event into non-Africans and back to Africa migrations, as similarly suggested to explain the finding of Neanderthal introgression sequences in the 1000 GP African populations or the presence of a putative modern human haplotype at the KNL1 gene in some Neanderthals [44,45]. The protein structure of ZnT9 predicted by AlphaFold [46] shows a total of 6 transmembrane domains and a cytosolic N-terminus for ZnT9, a structure similar to that predicted for the other members of the ZnT family. The location of the Met50Val substitution in the N-terminus may indicate a putative effect of this non-synonymous SNP on zinc homeostasis through a possible accessory role of the N-terminus domain of ZnT9 in facilitating zinc transport or protein interactions. When exploring the consequences of the Met50Val substitution at the molecular level, the protein expression analyses revealed a similar partial localization of both ZnT9 variants in the ER and mitochondria. Moreover, we observed mitochondrial clustering near the ER when ZnT9 was overexpressed what could imply a specific role in mitochondria-ER contact sites and architecture. The presence of the transporter in both cellular compartments has been previously described [47], and the ionic fluxes between the two organelles are highly interconnected. In this context, it has been recently described that the ZnT9KO phenotype can be rescued by targeting transporters in the ER and mitochondria [22]. The ER plays a major role in protein folding, maturation, quality control, and trafficking. Many of these processes require zinc to function correctly, and its deficit or excess often leads to protein malformations that induce ER stress. The most well-known ZTs implicated in ER zinc homeostasis are the ZnT5-ZnT6 heterodimer and ZnT7 homodimer, which are involved in zinc importation [48,49], and ZIP7, which is the main zinc-releasing transporter in the ER [50]. Considering the effect on ER zinc homeostasis of the overexpression of either ZnT9 isoform, we propose that the 50Val variant avoids ER zinc accumulation due to the reduced ZnT5-ZnT6 heterodimer expression. Additionally, we observed that ZIP7 inhibition generates an excess of zinc in the ER, supporting the idea that the zinc exporting activity of ZIP7 is key for maintaining endoplasmic zinc homeostasis. Therefore, it is likely that the observed absence of differences in ER zinc content in basal conditions is due to the master role of ZIP7 in fine-tuning ER zinc content. It has been proposed that ZnT9 in the mitochondria together with ZnT5 in the ER determine the zinc content in the mitochondrial matrix [22]. ZnT9 has been identified as a zinc exporter from the mitochondria based on the zinc accumulation observed in the mitochondria in ZnT9KO models [21,22,47]. We only detected a mild increase in zinc in ZnT9-silenced cells, which was not statistically significant. Probably, the zinc accumulation observed in KO models is the result of a complete and chronic effect. Nevertheless, this transport model implies that ZnT9 works in the opposite direction from the rest of the ZnT transport family. Our results showed that ZnT9 50Val expression increases the zinc content of the mitochondrial matrix but avoids zinc overload. The intermembrane space, however, showed no differences at high zinc concentrations between the tested conditions. Zinc homeostasis in this space has been previously reported to follow cytosolic zinc concentrations [51]. On the other hand, it is likely that the repression of ZnT5 expression observed in ZnT9 50Val cells might be a negative feedback to avoid a further zinc increase in the mitochondrial matrix. Having ruled out that ZnT9 is the main mitochondria zinc importer, we speculate that it might influence a wider machinery of proteins involved in mitochondrial zinc homeostasis. In this respect, ZnT9 has been recently connected to the expression of mitochondrial ribosomes and OxPhos components [52]. Importantly, we show that the ZnT9 50Val variant increases mitochondrial metabolism in a zinc-dependent manner. ZnT9 has been previously reported to affect mitochondrial function [21,22] and in this work we show that it also affects zinc homeostasis in the ER. Both organelles are interconnected and essential for cell fate and metabolism [53]. Previously, a mutation in the fourth TMD in SLC30A9 was found to cause acute dysregulation of zinc homeostasis in all family members affected by a novel cerebro-renal syndrome with important early neurological deterioration [54]. Given that the Met50Val substitution sits in what is expected to be an auxiliary region for the activity of the transporter, we hypothesize that the effects of rs1047626 on zinc homeostasis are milder, but somehow became adaptive in the past. Accordingly, when analyzing whether the genotype for rs1047626 influenced zinc content in the liver as a proxy for systemic zinc homeostasis, we only detected a tendency towards increased zinc in the liver when the derived G allele (50Val) was present. Even if larger sample sizes might produce results with significant differences, any putative effect of rs1047626 on liver zinc content would be lower than that of the QTLs previously identified in the 3´region of the SLC30A9 gene. Although SLC30A9 is moderately expressed in most tissues, its expression is higher in the fetal brain, cerebellum, skeletal muscle, pituitary, thyroid, and kidneys [54]. Hence, greater effects of the Met50Val substitution on other tissues or organs cannot be ruled out, for instance, related to adjustments in renal excretion, brain function or even skeletal muscle metabolism. The major haplotype outside Africa was found significantly associated with greater susceptibility to major depression and other related psychiatric disorders [55,56] as well as with higher values for self-reported helping behavior [57]. Moreover, our PheWAS analysis confirms that the derived G-allele of rs1047626 contributes not only to a greater risk for both major depressive disorder and neuroticism but also lower impedance (i.e., a higher lean body mass), among others. In the last years it is emerging the crucial function of ZnT9 in core mitochondrial processes. The function of the mitochondrial ribosomes (mitoribosomes) and, consequently, the expression of proteins encoded by mitochondrial DNA, such as several OxPhos subunits, has been shown to depend on ZnT9 expression [52]. In this context, our data proves that the Met50Val substitution influences mitochondrial metabolism. This effect might have an impact on systemic metabolism that could explain the association of rs1047626 with higher lean body mass. However, despite all the data we have compiled, it is not obvious how the observed differential molecular phenotype for the Met50Val substitution can be translated into an adaptive phenotype at the organism level or might relate to environmental zinc [13]. Given the essential role of the mitochondria in skeletal muscle thermogenesis, we propose that the derived G allele, which we have shown provides better protection to the mitochondria, may have been positively selected to facilitate adaptation to cold while providing higher susceptibility to neuropsychiatric traits. This is a plausible scenario for a case of archaic adaptive introgression, as Denisovans and Neanderthals were probably already well adapted to the local environmental conditions modern humans encountered when expanding across Eurasia. Some limitations in this study should be noted. Although several promising human traits have been associated with rs1047626, further work is required to understand how the molecular phenotype of differential zinc handling between the mitochondrial and the endoplasmic reticulum described here for the Met50Val substitution is refined into an adaptive phenotype at the organismal level. Similarly, we note that it cannot be ruled out the possible adaptive role of regulatory variants, influencing both SLC30A9 expression and liver zinc content (and thus systemic homeostasis of zinc), also contributing to the adaptive phenotype we proposed for the Met50Val substitution. After understanding where in the cell ZnT9 activity is having an impact and demonstrating that there are zinc cellular homeostasis differences between the Met50Val ZnT9 variants using an overexpression approach, other experimental assays working with endogenous expression ZnT9 levels might help to better characterize the putative disruption role of the Met50Val substitution in particular cell types and tissues of interest, such as muscle and adipose tissue or neuronal cells, among others. Finally, additional work is necessary to ascertain the exact timing, place, and potential admixture events experienced with archaic humans to explain the current patterns of variation and contrasting signatures of positive selection observed along the SLC30A9 gene region among modern human populations.
What the sources say
- Supportedcites [13] C Zhang, 2015, Differential natural selection of human zinc transporter genes between african and non-African populations., Sci Rep source
The Results section states that a previous study found SLC30A9 to be strongly selected in East Asian populations and ranked it among the top 10 genome-wide signals of positive selection. The Discussion likewise describes earlier reports of a strong selection signal in this gene. (LLM, conf=0.94)
- Not checked — no full text of the sourcecites [14] LB Barreiro, 2008, Natural selection has driven population differentiation in modern humans, Nat Genet source
- Could not verifycites [16] CS Carlson, 2005, Genomic regions exhibiting positive selection identified from dense genotype data, Genome Res sourceExperimental — accuracy not yet measured
The Results and Screening for positive selection sections report candidate regions with signatures of selection, but the passages found do not identify SLC30A9 or link its region to one of those signals. Searches for SLC30A9 and zinc-transporter terminology found no passage addressing the specific claim. (LLM, conf=0.76)
- Could not verifycites [17] SR Grossman, 2013, Identifying Recent Adaptations in Large-Scale Genomic Data, Cell sourceExperimental — accuracy not yet measured
The Results section (‘412 Fine-mapped Signals of Selection’; page unavailable) reports candidate regions under selection, but the retrieved passages do not identify SLC30A9 as one of them. Searches for SLC30A9 and zinc-transporter terminology found no passage connecting that region to a selection signal. (LLM, conf=0.73)
- Supportedcites [20] SH Williamson, 2007, Localizing Recent Adaptive Evolution in the Human Genome., PLoS Genet source
Table 1 (Discussion; page not specified) lists SLC30A9 in the Chinese sample among the 101 regions with the strongest evidence for a recent selective sweep (CLR p < 0.00001). This supports the claim of a strong positive-selection signature in the SLC30A9 region, specifically in that sample. (LLM, conf=0.92)
- Not checked — no full text of the sourcecites [25] PC Sabeti, 2006, Positive Natural Selection in the Human Lineage, Science source
boostedstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
89 more claims · 3 with their cited sources checked
Create a free account to see all 141 claims and every source verdict.
- It may state two separate claims as one. 34
- Probably not a substantial claim. 11
- The quote only partly states this claim. 2
Citations
Cites
Highly accurate protein structure prediction with AlphaFold
Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Žídek, A., Potapenko, A., et al. (2021) · Nature · doi:10.1038/s41586-021-03819-2 · Cited by 43,764
A global reference for human genetic variation
The 1000 Genomes Project Consortium, authors, C., Auton, A., Abecasis, G. R., Steering committee, Altshuler, D. M., Durbin, R. M., Abecasis, G. R., Bentley, D. R., Chakravarti, A., et al. (2015) · Nature · doi:10.1038/nature15393 · Cited by 18,247
ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data
Wang, K., Li, M., & Hakonarson, H. (2010) · Nucleic Acids Research · doi:10.1093/nar/gkq603 · Cited by 14,621
A general framework for estimating the relative pathogenicity of human genetic variants
Kircher, M., Witten, D. M., Jain, P., O'Roak, B. J., Cooper, G. M., & Shendure, J. (2014) · Nature Genetics · doi:10.1038/ng.2892 · Cited by 6,183
Buniello, A., MacArthur, J. A. L., Cerezo, M., Harris, L. W., Hayhurst, J., Malangone, C., McMahon, A., Morales, J., Mountjoy, E., Sollis, E., et al. (2019) · Nucleic Acids Research · doi:10.1093/nar/gky1120 · Cited by 4,000 · cited 1×
“We used the GWAS catalog (https://www.ebi.ac.uk/gwas/) [69] and the PheWAS option in the GWAS atlas (https://atlas.ctglab.nl/) [70] to look for trait associations with rs1047626 and SNPs in high LD (r 2 > 0.8).”
InfluentialeQTLGen, Wray, N. R., 23andMe, the Major Depressive Disorder Working Group of the Psychiatric Genomics Consortium, Ripke, S., Mattheisen, M., Trzaskowski, M., Byrne, E. M., Abdellaoui, A., Adams, M. J., et al. (2018) · Nature Genetics · doi:10.1038/s41588-018-0090-3 · Cited by 3,005
The biochemical basis of zinc physiology
Vallee, B. L., & Falchuk, K. H. (1993) · Physiological Reviews · doi:10.1152/physrev.1993.73.1.79 · Cited by 2,990 · cited 1×
“Zinc is an essential micronutrient with different structural, catalytic, and regulatory roles in the human body [6–8].”
pegas: an R package for population genetics with an integrated–modular approach
Paradis, E. (2010) · Bioinformatics · doi:10.1093/bioinformatics/btp696 · Cited by 2,582
The complete genome sequence of a Neanderthal from the Altai Mountains
Prüfer, K., Racimo, F., Patterson, N., Jay, F., Sankararaman, S., Sawyer, S., Heinze, A., Renaud, G., Sudmant, P. H., de Filippo, C., et al. (2014) · Nature · doi:10.1038/nature12886 · Cited by 2,254
Genome-wide detection and characterization of positive selection in human populations
The International HapMap Consortium, Sabeti, P. C., Varilly, P., Fry, B., Lohmueller, J., Hostetter, E., Cotsapas, C., Xie, X., Byrne, E. H., McCarroll, S. A., et al. (2007) · Nature · doi:10.1038/nature06250 · Cited by 2,131
Research works printed but not linked: 1
- A Auton, 2015, A global reference for human genetic variation, Nature, 526, 68, 10.1038/nature15393 · DOI · Paper page
Cited by
Paoli‐Iseppi, R. D., Joshi, S. S., Gleeson, J., Prawer, Y. D. J., You, Y., Agarwal, R., Li, A., Hull, A., Whitehead, E. M., Seo, Y., et al. (2025) · Genome biology · doi:10.1186/s13059-025-03724-1 · Cited by 17
“FLAMES exhibited 100% recall for SIRV5 across all annotations, however, its performance with SIRV6 was suboptimal, indicating a higher degree of variability in the FLAMES isoform discovery pipeline.”
Comprehensive Roles of ZIP and ZnT Zinc Transporters in Metabolic Inflammation
Barman, S., Pradeep, S. R., & Srinivasan, K. (2026) · Targets · doi:10.3390/targets4010005 · Cited by 9
“ZIP transporters deliver cytosolic zinc for cellular processes, whereas ZnT transporters export excess zinc to ensure optimal storage for optimal cellular performance [12].”
Paoli‐Iseppi, R. D., Joshi, S. S., Gleeson, J., Prawer, Y. D. J., You, Y., Agarwal, R., Li, A., Hull, A., Whitehead, E. M., Seo, Y., et al. (2024) · medRxiv · doi:10.1101/2024.02.22.24303189 · Cited by 5
Martínez-Morata, I., Dennehy, K., Chepurny, O. G., Delibaşı, T., Schilling, K., Navas‐Acién, A., & Muayed, M. E. (2025) · Journal of Trace Elements in Medicine and Biology · doi:10.1016/j.jtemb.2025.127733 · Cited by 7
Genetic determinants of zinc homeostasis and its role in cardiometabolic diseases
Sadler, M. C., Ghobril, J.-P., Borisov, O., Perrais, M., Schiano, G., Petrović, D., Ha, E., Ponte, B., Li, Y., Rodriguez-Hernandez, Z., et al. (2025) · PLoS Genetics · doi:10.1371/journal.pgen.1011928 · Cited by 4
Metadata
- Authors
- Roca-Umbert, A., Garcia-Calleja, J., Vogel-González, M., Fierro-Villegas, A., Ill-Raga, G., Herrera-Fernández, V., Bosnjak, A., Muntané, G., Gutiérrez, E., Campelo, F., Vicente, R., & Bosch, E.
- Affiliations
- Universitat Pompeu Fabra (ES)
- Institute of Photonic Sciences (ES)
- Corresponding author
- Vicente, Rubén
- Bosch, Elena
- ORCID
- Venue
- PLoS Genetics
- Publisher
- Public Library of Science (PLoS)
- Published
- 2023-09-25
- Volume
- 19
- Issue
- 9
- Pages
- e1010950
- ISSN
- 1553-7390 (ISSN-L)
- 1553-7404
- PMC
- PMC10553801
- OpenAlex
- W4387011928
- Cited by
- 8
- References
- 74
- Citations by year
- 2026: 2 · 2025: 5 · 2024: 1
- FWCI
- 1.40
- Type
- journal article
- Language
- en
- Update notices
- new_version (2023-10-05)
- Open access
- gold
- doi.org
- Licence
- Locations
- PLoS Genetics (publishedVersion, cc-by)
- bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- Repositori digital de la UPF (Universitat Pompeu Fabra) (acceptedVersion, cc-by)
- Rovira & Virgili repository (submittedVersion, cc-by)
- DIGITAL.CSIC (Spanish National Research Council (CSIC)) (submittedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- Ministerio de Ciencia e Innovación (RYC-2017-22227, PID2019-106232RB-I00, PID2019-110933GB-I00, PID2019-106755RB-I00)
- Unidad de Excelencia María de Maeztu (CEX2018-000792-M)
- Severo Ochoa (CEX2019-000910-S)
- Direcció General de Recerca, Generalitat de Catalunya (2017SGR00702)
- Fundació Cellex, Fundació Mir-Puig, and Generalitat de Catalunya
- FPI-MCIN/AEI PhD (PRE2020-095762)
- Instituto de Salud Carlos III (PI21/00612)
- Topics
- Health Sciences › Nursing › Nutrition and Dietetics › Trace Elements in Health
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Associations and Epidemiology
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Forensic and Genetic Research
- Keywords
- zinc homeostasis, positive selection, directional selection, Denisovan introgression, gain-of-function, mitochondrial metabolism, cold adaptation, neuropsychiatric disorders, skeletal muscle
- MeSH terms
- Human Genetics, Animals, Haplotypes, Homeostasis, Humans, Selection, Genetic, Zinc, Hominidae, Genome, Human, HEK293 Cells
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-11, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-11 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol and qwen3.8-omni-flash).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Human genetic adaptation related to cellular zinc homeostasis. https://app.cochat.ai/papers/10.1371/journal.pgen.1010950
- The paper
- Roca-Umbert, A., Garcia-Calleja, J., Vogel-González, M., Fierro-Villegas, A., Ill-Raga, G., Herrera-Fernández, V., Bosnjak, A., Muntané, G., Gutiérrez, E., Campelo, F., Vicente, R., & Bosch, E. (2023). Human genetic adaptation related to cellular zinc homeostasis. PLoS Genetics. https://doi.org/10.1371/journal.pgen.1010950
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.