Probabilistic ancestry maps: a method to assess and visualize population substructures in genetics

Estimation of genetic variation in vitiligo associated genes: Population genomics perspective

BACKGROUND

Vitiligo is an auto-immune progressive depigmentation disorder of the skin due to loss of melanocytes. Genetic risk is one of the important factors for development of vitiligo. Preponderance of vitiligo in certain ethnicities is known which can be analysed by understanding the distribution of allele frequencies across normal populations. Earlier GWAS identified 108 risk alleles for vitiligo in Europeans and East Asians. In this study, 64 of these risk alleles were used for analysing their enrichment and depletion across populations (1000 Genomes Project and IndiGen) with reference to 1000 Genomes dataset. Genetic risk scores were calculated and Fisher's exact test was performed to understand statistical significance of their variation in each population with respect to 1000 Genomes dataset as reference. In addition to SNPs reported in GWAS, significant variation in allele frequencies of 1079 vitiligo-related genes were also analysed. Two-tailed Chi-square test and Bonferroni's multiple adjustment values along with fixation index (≥ 0.5) and minimum allele frequency (≥ 0.05) were calculated and used to prioritise the variants based on pairwise comparison across populations.

RESULTS

Risk alleles rs1043101 and rs10768122 belong to 3 prime UTR of glutamate receptor gene SLC1A2 are found to be highly enriched in the South Asian population when compared with the 'global normal' population. Intron variant rs4766578 (ATXN2) was found to be deleted in SAS, EAS and AFR and enriched in EUR and AMR1. This risk allele is found to be under positive selection in SAS, AMR1 and EUR. From the ancillary vitiligo gene list, nonsynonymous variant rs16891982 was found to be enriched in the European and the Admixed American populations and depleted in all others. rs2279238 and rs11039155 belonging to the LXR-α gene involved in regulation of metalloproteinase 2 and 9 (melanocyte precursors) were found to be associated with vitiligo in the North Indian population (in earlier study).

CONCLUSION

The differential enrichment/depletion profile of the risk alleles provides insight into the underlying inter-population variations. This would provide clues towards prioritisation of SNPs associated with vitiligo thereby elucidating its preponderance in different ethnic groups.

Predicting S. aureus antimicrobial resistance with interpretable genomic space maps

Increasing antimicrobial resistance (AMR) represents a global healthcare threat. To decrease the spread of AMR and associated mortality, methods for rapid selection of optimal antibiotic treatment are urgently needed. Machine learning (ML) models based on genomic data to predict resistant phenotypes can serve as a fast screening tool prior to phenotypic testing. Nonetheless, many existing ML methods lack interpretability. Therefore, we present a methodology for visualization of sequence space and AMR prediction based on the non-linear dimensionality reduction method - generative topographic mapping (GTM). This approach, applied to AMR data of >5000 S. aureus isolates retrieved from the PATRIC database, yielded GTM models with reasonable accuracy for all drugs (balanced accuracy values ≥0.75). The Generative Topographic Maps (GTMs) represent data in the form of illustrative maps of the genomic space and allow for antibiotic-wise comparison of resistant phenotypes. The maps were also found to be useful for the analysis of genetic determinants responsible for drug resistance. Overall, the GTM-based methodology is a useful tool for both the illustrative exploration of the genomic sequence space and AMR prediction.

Prevalence, Morbidity, and Mortality of Men With Sex Chromosome Aneuploidy in the Million Veteran Program Cohort

Importance

The reported phenotypes of men with 47,XXY and 47,XYY syndromes include tall stature, multisystem comorbidities, and poor health-related quality of life (HRQOL). However, knowledge about these sex chromosome aneuploidy (SCA) conditions has been derived from studies in the less than 15% of patients who are clinically diagnosed and also lack diversity in age and genetic ancestry.

Objectives

To determine the prevalence of clinically diagnosed and undiagnosed X or Y chromosome aneuploidy among men enrolled in the Million Veteran Program (MVP); to describe military service metrics of men with SCAs; and to compare morbidity and mortality outcomes between men with SCA with and without a clinical diagnosis vs matched controls.

Design, Setting, and Participants

This cross-sectional study used a case-control recruitment design to select biological males enrolled in the MVP biobank in the US Veterans Administration health care system from 2011 to 2022. Cases were participants with 47,XXY syndrome or 47,XYY syndrome, matched 1:5 with controls based on sex, age, and genetic ancestry. Data were analyzed from January 2022 to December 2023.

Exposure

Genomic identification of an additional X or Y chromosome.

Main Outcomes and Measures

Outcomes of interest included prevalence of men with SCAs from genomic analysis; clinical SCA diagnosis; Charlson Comorbidity Index; rates of outpatient, inpatient, and emergency encounters per year; self-reported health outcomes; and standardized mortality ratio.

Results

Of 595 612 genotyped males in the MVP, 862 had an additional X chromosome (47,XXY) and 747 had an extra Y chromosome (47,XYY), with the highest prevalence among men with East Asian (47,XXY: 10 of 7313 participants; 47,XYY: 14 of 7313 participants) and European (47,XXY: 725 of 427 143 participants; 47,XYY: 625 of 427 143 participants) ancestry. Mean (SD) age at assessment was 61 (12) years, at which point 636 veterans (74.X%) with 47,XXY and 745 veterans (99%) with 47,XYY remained undiagnosed. Individuals with 47,XXY and 47,XYY had similar military service history, all-cause standardized mortality ratio, and age of death compared with matched controls. Individuals with SCA, compared with controls, had higher Charlson Comorbidity Index scores (47,XXY: mean [SD], 4.30 [2.72] vs controls: mean [SD], 3.90 [2.47]; 47,XYY: mean [SD], 4.45 [2.90] vs controls: mean [SD], 3.82 [2.50]) and health care utilization (eg, median [IQR] outpatient encounters per year: 47,XXY, 22.6 [11.8-37.8] vs controls, 16.8 [9.4-28]; 47,XYY: 21.4 [12.4-33.8] vs controls: 17.0 [9.4-28.2]), while several measures of HRQOL were lower (eg, mean [SD] self-reported physical function: 47,XXY: 34.2 [12] vs control mean [SD] 37.8 [12.8]; 47,XYY: 36.3 [11.6] vs control 37.9 [12.8]). Men with a clinical diagnosis of 47,XXY, compared with individuals without a clinical diagnosis, had higher health care utilization (eg, median [IQR] encounters per year: 26.6 [14.9-43.2] vs 22.2 [11.3-36.0]) but lower Charlson Comorbidity Index scores (mean [SD]: 3.7 [2.7] vs 4.5 [4.1]).

Conclusion and Relevance

In this case-control study of men with 47,XXY and 47,XYY syndromes, prevalence of SCA was comparable with estimates in the general population. While these men had successfully served in the military, they had higher morbidity and reported poorer HRQOL with aging. Longer longitudinal follow-up of this sample will be informative for clinical and patient-reported outcomes, the role of ancestry, and mortality statistics.

Analysis of genetic diversity in patients with major psychiatric disorders versus healthy controls: A molecular-genetic study of 1698 subjects genotyped for 100 candidate genes (549 SNPs)

BACKGROUND

This study analyzed the extent to which irregularities in genetic diversity separate psychiatric patients from healthy controls.

METHODS

Genetic diversity was quantified through multidimensional "gene vectors" assembled from 4 to 8 polymorphic SNPs located within each of 100 candidate genes. The number of different genotypic patterns observed per gene was called the gene's "diversity index".

RESULTS

The diversity indices were found to be only weakly correlated with their constituent number of SNPs (20.5 % explained variance), thus suggesting that genetic diversity is an intrinsic gene property that has evolved over the course of evolution. Significant deviations from "normal" diversity values were found for (1) major depression; (2) Alzheimer's disease; and (3) schizoaffective disorders. Almost one third of the genes were correlated with each other, with correlations ranging from 0.0303 to 0.7245. The central finding of this study was the discovery of "singular genes" characterized by distinctive genotypic patterns that appeared exclusively in patients but not in healthy controls. Neural Nets yielded nonlinear classifiers that correctly identified up to 90 % of patients. Overlaps between diagnostic subgroups on the genotype level suggested that (1) diagnoses-crossing vulnerabilities are likely involved in the pathogenesis of major psychiatric disorders; (2) clinically defined diagnoses may not constitute etiological entities.

CONCLUSION

Detailed analyses of the variation of genotypic patterns in genes along with the correlation between genes lead to nonlinear classifiers that enable very robust separation between psychiatric patients and healthy controls on the genotype level.

Operationalisation of the ForenSeq® Kintelligence Kit for Australian unidentified and missing persons casework

Single nucleotide polymorphism (SNP) genotyping technologies can generate investigative leads for human remains identification, including estimation of biological sex, biogeographical ancestry (BGA), externally visible characteristics (EVCs), identity, uniparental lineage and extended kinship. The ForenSeq® Kintelligence Kit provides forensic laboratories with the ability to apply this suite of genetic tools to forensic samples using one panel targeting 10,230 SNPs (including 56 ancestry-informative, 24 phenotype-informative, 94 identity-informative, 106 X chromosome, 85 Y chromosome and 9867 kinship-informative SNPs) sequenced on the MiSeq FGx® Sequencing System. The ForenSeq® Kintelligence Kit has been internally validated, optimised and operationalised by the Australian Federal Police National DNA Program for Unidentified and Missing Persons (AFP Program) for coronial casework. The internal validation was conducted according to the Scientific Working Group on DNA Analysis Methods guidelines (excluding mixture analysis), focussing on sample types typically encountered in human remains identification casework, such as bones, teeth, nail, blood and hair. The workflow was optimised for a high throughput library preparation and sequencing workflow, and additional analytical thresholds were developed to improve genotyping accuracy for low DNA input samples. Additionally, the genetic intelligence generated from the kit was compared to the self-declared biological sex, EVCs and BGA of the DNA donors to assess concordance. The kit was able to produce high quality SNP profiles from 1.0 ng down to 0.1 ng of DNA, with high repeatability and reproducibility, and minimal background noise. The prediction accuracy for biological sex (95%), hair colour (58%), eye colour (74%) and BGA inferences (consistent: 74%; partially consistent: 10%; inconclusive: 16%) was determined based on self-declared data. Additionally, SNP profiles from a volunteer family group of ten related individuals were uploaded to GEDmatch PRO™ to assess kinship accuracy. The kit was capable of detecting (97%) and accurately classifying (90%) genetic relationships spanning from first to fifth degree. The Kintelligence Kit provides the AFP Program with a robust and reliable genetic intelligence tool for unidentified and missing persons investigations, which has been designed to sequence multiple challenging samples in a single multiplexed assay using existing laboratory instrumentation.

Adherence to a healthy lifestyle, genetic susceptibility to abdominal obesity, cardiometabolic risk markers, and risk of coronary heart disease

BACKGROUND

Little is known about whether the association between genetic susceptibility to high waist-to-hip ratio (WHR), a measure of abdominal obesity, and incident coronary heart disease (CHD) is modified by adherence to a healthy lifestyle.

OBJECTIVES

To explore the interplay of genetic susceptibility to high WHR and adherence to a healthy lifestyle on incident CHD.

METHODS

This study included 282,316 white British individuals from the UK Biobank study. Genetic risk for high WHR was estimated in the form of weighted polygenic risk scores (PRSs), calculated based on 156 single-nucleotide polymorphisms. Lifestyle scores were calculated based on 5 healthy lifestyle factors: regular physical activity, no current smoking, a healthy diet, <3 times/wk of alcohol consumption and 7-9 h/d of sleep. Incident CHD (n = 11,635) was accrued over a median 13.8 y of follow-up, and 12 individual cardiovascular disease risk markers assessed at baseline.

RESULTS

Adhering to a favorable lifestyle (4-5 healthy factors) was associated with a 25% (hazard ratio: 0.75, 95% confidence interval: 0.70, 0.81) lower hazard of CHD compared with an unfavorable lifestyle (0-1 factor), independent of PRS for high WHR. Estimated 12-y absolute risk of CHD was lower for a favorable lifestyle at high genetic risk (1.73%) and medium genetic risk (1.67%) than for an unfavorable lifestyle at low genetic risk (2.08%). Adhering to a favorable lifestyle was associated with healthier levels of cardiovascular disease risk markers (except random glucose and high-density lipoprotein), independent of PRS for high WHR.

CONCLUSIONS

Individuals who have high or medium genetic risk of abdominal obesity but adhere to a healthy lifestyle may have a lower risk of developing CHD, compared with those who have low genetic risk and an unhealthy lifestyle. Future clinical trials of lifestyle modification could be implemented for individuals at high genetic risk of abdominal obesity for the primary prevention of CHD events.

Ethnicity- and sex-specific genome wide association study on Parkinson's disease

Most previous genome-wide association studies (GWASs) on Parkinson's disease (PD) focus on the European population. There are several sex-specific clinical differences in PD, but little is known about its genetic background. We aimed to perform an ethnicity-specific and sex-specific GWAS on PD in the Korean population. A total of 1050 PD patients and 5000 controls were included. For primary analysis, we performed a GWAS using a logistic additive model adjusted for age and sex. The same statistical models were applied to sex-specific analyses. Genotyping was performed using a customized microarray chip optimized for the Korean population. Nine single nucleotide polymorphisms (SNPs) including four in the SNCA locus and three from the PARK16 locus were associated with PD in Koreans. The rs34778348 in the LRRK2 locus showed a strong association, though failed to pass cluster quality control. There were no notable genome-wide significant markers near the MAPT or GBA1 loci. In the female-only analysis, rs34778348 in LRRK2 and the four other SNPs in the SNCA showed a strong association with PD. In the male-only analysis, no SNP surpassed the genome-wide significance threshold under Bonferroni correction; however, the most significant signal was rs708726 in the PARK16 locus. This ethnicity- and sex-specific GWAS on PD implicate the pan-ethnic effect of SNCA, the universal but East-Asian inclined effect of PARK16, the East Asian-specific role of LRRK2 G2385R variants, and the possible disproportionate effect of SNCA and PARK16 between sexes for PD susceptibility. These findings suggest the different genetic contributions to sporadic PD in terms of ethnicity and sex.

Prevalence, Morbidity, and Mortality of 1,609 Men with Sex Chromosome Aneuploidy: Results from the Diverse Million Veteran Program Cohort

Importance

The reported phenotypes of men with 47,XXY and 47,XYY syndromes include tall stature, multisystem comorbidities, and poor health-related quality of life (HRQoL). However, knowledge about these sex chromosome aneuploidy (SCA) conditions has been derived from studies in the <15% of patients who are clinically diagnosed and also lack diversity in age and genetic ancestry.

Objectives

Determine the prevalence of clinically diagnosed and undiagnosed X or Y chromosome aneuploidy among men enrolled in the Million Veteran Program (MVP); describe military service metrics of men with SCAs; compare morbidity and mortality outcomes between men with SCA with and without a clinical diagnosis to matched controls.

Design

Cross-sectional, case-control.

Setting

United States Veterans Administration Healthcare System.

Participants

Biologic males enrolled in the MVP biobank with genomic identification of an additional X or Y chromosome (cases); controls matched 1:5 on sex, age, and genetic ancestry.

Main Outcomes and Measures

Prevalence of men with SCAs from genomic analysis; clinical SCA diagnosis; Charlson Comorbidity Index (CCI); rates of outpatient, inpatient, and emergency encounters per year; self-reported health outcomes; standardized mortality ratio (SMR).

Results

An additional X or Y chromosome was present in 145 and 125 per 100,000 males in the MVP, respectively, with the highest prevalence among men with European and East Asian ancestry. At a mean age of 61±12 years, 74% of male veterans with 47,XXY and >99% with 47,XYY remained undiagnosed. Individuals with 47,XXY (n=862) and 47,XYY (n=747) had similar military service history, all-cause SMR, and age of death compared to matched controls. CCI and healthcare utilization were higher among individuals with SCA, while several measures of HRQoL were lower. Men with a clinical diagnosis of 47,XXY had higher healthcare utilization but lower comorbidity score compared to those undiagnosed.

Conclusion and Relevance

One in 370 males in the MVP cohort have SCA, a prevalence comparable to estimates in the general population. While these men have successfully served in the military, they have higher morbidity and report poorer HRQoL with aging. Longer longitudinal follow-up of this sample will be informative for clinical and patient-reported outcomes, the role of ancestry, and mortality statistics.

KEY POINTS

Comparable to the general population, approximately 1 in 370 male veterans have a sex chromosome aneuploidy, but most are undiagnosed.Men with X or Y chromosome aneuploidy successfully complete US miliary duty with similar service history compared to their 46,XY peers.Medical comorbidities and healthcare utilization metrics are higher in male veterans with 47,XXY and 47,XYY during aging, however life expectancy is similar to matched controls.

Case-control association study of congenital heart disease from a tertiary paediatric cardiac centre from North India

BACKGROUND

Congenital Heart diseases (CHDs) account for 1/3rd of all congenital birth defects. Etiopathogenesis of CHDs remain elusive despite extensive investigations globally. Phenotypic heterogeneity witnessed in this developmental disorder reiterate gene-environment interactions with periconceptional factors as risk conferring; and genetic analysis of both sporadic and familial forms of CHD suggest its multigenic basis. Significant association of de novo and inherited variants have been observed. Approximately 1/5th of CHDs are documented in the ethnically distinct Indian population but genetic insights have been very limited. This pilot case-control based association study was undertaken to investigate the status of Caucasian SNPs in a north Indian cohort.

METHOD

A total of 306 CHD cases sub-classified into n = 198 acyanotic and n = 108 cyanotic types were recruited from a dedicated tertiary paediatric cardiac centre in Palwal, Haryana. 23 SNPs primarily prioritized from Genome-wide association studies (GWAS) on Caucasians were genotyped using Agena MassARRAY Technology and test of association was performed with adequately numbered controls.

RESULTS

Fifty percent of the studied SNPs were substantially associated in either allelic, genotypic or sub-phenotype categories validating their strong correlation with disease manifestation. Of note, strongest allelic association was observed for rs73118372 in CRELD1 (p < 0.0001) on Chr3, rs28711516 in MYH6 (p = 0.00083) and rs735712 in MYH7 (p = 0.0009) both on Chr 14 and were also significantly associated with acyanotic, and cyanotic categories separately. rs28711516 (p = 0.003) and rs735712 (p = 0.002) also showed genotypic association. Strongest association was observed with rs735712(p = 0.003) in VSD and maximum association was observed for ASD sub-phenotypes.

CONCLUSIONS

Caucasian findings were partly replicated in the north Indian population. The findings suggest the contribution of genetic, environmental and sociodemographic factors, warranting continued investigations in this study population.

Biogeographical ancestry, variable selection, and PLS-DA method: a new panel to assess ancestry in forensic samples via MPS technology

As evidenced by the large number of articles recently published in the literature, forensic scientists are making great efforts to infer externally visible features and biogeographical ancestry (BGA) from DNA analysis. Just as phenotypic, ancestry information obtained from DNA can provide investigative leads to identify the victims (missing/unidentified persons, crime/armed conflict/mass disaster victims) or trace their perpetrators when no matches were found with the reference profile or in the database. Recently, the advent of Massively Parallel Sequencing technologies associated with the possibility of harnessing high-throughput genetic data allowed us to investigate the associations between phenotypic and genomic variations in worldwide human populations and develop new BGA forensic tools capable of simultaneously analyzing up to millions of markers if for example the ancient DNA approach of hybridization capture was adopted to target SNPs of interest. In the present study, a selection of more than 3000 SNPs was performed to create a new BGA panel and the accuracy of the new panel to infer ancestry from unknown samples was evaluated by the PLS-DA method. Subsequently, the panel created was assessed using three variable selection techniques (Backward variable elimination, Genetic Algorithm and Regularized elimination procedure), and the best SNPs in terms of inferring bio-geographical ancestry at inter- and intra-continental level were selected to obtain panels to predict BGA with a reduced number of selected markers to be applied in routine forensic cases where PCR amplification is the best choice to target SNPs.

A new 165-SNP low-density lipoprotein cholesterol polygenic risk score based on next generation sequencing outperforms previously published scores in routine diagnostics of familial hypercholesterolemia

Genetic diagnosis of familial hypercholesterolemia (FH) remains unexplained in 30 to 70% of patients after exclusion of monogenic disease. There is now a growing evidence that a polygenic burden significantly modulates LDL-cholesterol (LDL-c) concentrations. Several LDL-c polygenic risk scores (PRS) have been set up. However, the balance between their diagnosis performance and their practical use in routine practice is not clearly established. Consequently, we set up new PRS based on our routine panel for sequencing and compared their diagnostic performance with previously-published PRS. After a meta-analysis, four new PRS including 165 to 1633 SNP were setup using different softwares. They were established using two French control cohorts (MONA LISA n=1082 and FranceGenRef n=856). Then the explained LDL-c variance and the ability of each PRS to discriminate monogenic negative FH patients (M-) versus healthy controls were compared with 4 previously-described PRS in 785 unrelated FH patients. Between all PRS, the 165-SNP PRS developed with PLINK showed the best LDL-c explained variance (adjusted R²=0.19) and the best diagnosis abilities (AUROC=0.77, 95%CI=0.74-0.79): it significantly outperformed all the previously-published PRS (p<1 × 10^-4). By using a cut-off at the 75th percentile, 61% of M- patients exhibited a polygenic hypercholesterolemia with the 165-SNP PRS versus 48% with the previously published 12-SNP PRS (p =3.3 × 10^-6). These results were replicated using the UK biobank. This new 165-SNP PRS, usable in routine diagnosis, exhibits better diagnosis abilities for a polygenic hypercholesterolemia diagnosis. It would be a valuable tool to optimize referral for whole genome sequencing.

Longitudinal cardio-respiratory fitness prediction through wearables in free-living environments

Cardiorespiratory fitness is an established predictor of metabolic disease and mortality. Fitness is directly measured as maximal oxygen consumption (VO2max), or indirectly assessed using heart rate responses to standard exercise tests. However, such testing is costly and burdensome because it requires specialized equipment such as treadmills and oxygen masks, limiting its utility. Modern wearables capture dynamic real-world data which could improve fitness prediction. In this work, we design algorithms and models that convert raw wearable sensor data into cardiorespiratory fitness estimates. We validate these estimates' ability to capture fitness profiles in free-living conditions using the Fenland Study (N=11,059), along with its longitudinal cohort (N = 2675), and a third external cohort using the UK Biobank Validation Study (N = 181) who underwent maximal VO2max testing, the gold standard measurement of fitness. Our results show that the combination of wearables and other biomarkers as inputs to neural networks yields a strong correlation to ground truth in a holdout sample (r = 0.82, 95CI 0.80-0.83), outperforming other approaches and models and detects fitness change over time (e.g., after 7 years). We also show how the model's latent space can be used for fitness-aware patient subtyping paving the way to scalable interventions and personalized trial recruitment. These results demonstrate the value of wearables for fitness estimation that today can be measured only with laboratory tests.

Maternal iron status in early pregnancy and DNA methylation in offspring: an epigenome-wide meta-analysis

BACKGROUND

Unbalanced iron homeostasis in pregnancy is associated with an increased risk of adverse birth and childhood health outcomes. DNA methylation has been suggested as a potential underlying mechanism linking environmental exposures such as micronutrient status during pregnancy with offspring health. We performed a meta-analysis on the association of maternal early-pregnancy serum ferritin concentrations, as a marker of body iron stores, and cord blood DNA methylation. We included 1286 mother-newborn pairs from two population-based prospective cohorts. Serum ferritin concentrations were measured in early pregnancy. DNA methylation was measured with the Infinium HumanMethylation450 BeadChip (Illumina). We examined epigenome-wide associations of maternal early-pregnancy serum ferritin and cord blood DNA methylation using robust linear regression analyses, with adjustment for confounders and performed fixed-effects meta-analyses. We additionally examined whether associations of any CpGs identified in cord blood persisted in the peripheral blood of older children and explored associations with other markers of maternal iron status. We also examined whether similar findings were present in the association of cord blood serum ferritin concentrations with cord blood DNA methylation.

RESULTS

Maternal early-pregnancy serum ferritin concentrations were inversely associated with DNA methylation at two CpGs (cg02806645 and cg06322988) in PRR23A and one CpG (cg04468817) in PRSS22. Associations at two of these CpG sites persisted at each of the follow-up time points in childhood. Cord blood serum ferritin concentrations were not associated with cord blood DNA methylation levels at the three identified CpGs.

CONCLUSION

Maternal early-pregnancy serum ferritin concentrations were associated with lower cord blood DNA methylation levels at three CpGs and these associations partly persisted in older children. Further studies are needed to uncover the role of these CpGs in the underlying mechanisms of the associations of maternal iron status and offspring health outcomes.

A deep learning framework for characterization of genotype data

Dimensionality reduction is a data transformation technique widely used in various fields of genomics research. The application of dimensionality reduction to genotype data is known to capture genetic similarity between individuals, and is used for visualization of genetic variation, identification of population structure as well as ancestry mapping. Among frequently used methods are principal component analysis, which is a linear transform that often misses more fine-scale structures, and neighbor-graph based methods which focus on local relationships rather than large-scale patterns. Deep learning models are a type of nonlinear machine learning method in which the features used in data transformation are decided by the model in a data-driven manner, rather than by the researcher, and have been shown to present a promising alternative to traditional statistical methods for various applications in omics research. In this study, we propose a deep learning model based on a convolutional autoencoder architecture for dimensionality reduction of genotype data. Using a highly diverse cohort of human samples, we demonstrate that the model can identify population clusters and provide richer visual information in comparison to principal component analysis, while preserving global geometry to a higher extent than t-SNE and UMAP, yielding results that are comparable to an alternative deep learning approach based on variational autoencoders. We also discuss the use of the methodology for more general characterization of genotype data, showing that it preserves spatial properties in the form of decay of linkage disequilibrium with distance along the genome and demonstrating its use as a genetic clustering method, comparing results to the ADMIXTURE software frequently used in population genetic studies.

Exploring gene-culture coevolution in humans by inferring neuroendophenotypes: A case study of the oxytocin receptor gene and cultural tightness

The gene-culture coevolution (GCC) framework has gained increasing prominence in the social and biological sciences. While most studies on human GCC concern the evolution of low-level physiological traits, attempts have also been made to apply GCC to complex human traits, including social behavior and cognition. One major methodological challenge in this endeavor is to reconstruct a specific biological pathway between the implicated genes and their distal phenotypes. Here, we introduce a novel approach that combines data on population genetics and expression quantitative trait loci to infer the specific intermediate phenotypes of genes in the brain. We suggest that such "neuroendophenotypes" will provide more detailed mechanistic insights into the GCC process. We present a case study where we explored a GCC dynamics between the oxytocin receptor gene (OXTR) and cultural tightness-looseness. By combining data from the 1000 Genomes project and the Gene-Tissue-Expression project (GTEx), we estimated and compared OXTR expression in 10 brain regions across five human superpopulations. We found that OXTR expression in the anterior cingulate cortex (ACC) was highly variable across populations, and this variation correlated with cultural tightness and socio-ecological threats worldwide. The mediation models also suggested possible GCC dynamics where the increased OXTR expression in the ACC mediates or emerges from the tight culture and higher socio-ecological threats. Formal selection scans further confirmed that OXTR alleles linked to enhanced receptor expression in the ACC underwent positive selection in East Asian countries with tighter social norms. We discuss the implications of our method in human GCC research.

Inference of recent admixture using genotype data

The inference of biogeographic ancestry (BGA) has become a focus of forensic genetics. Misinference of BGA can have profound unwanted consequences for investigations and society. We show that recent admixture can lead to misclassification and erroneous inference of ancestry proportions, using state of the art analysis tools with (i) simulations, (ii) 1000 genomes project data, and (iii) two individuals analyzed using the ForenSeq DNA Signature Prep Kit. Subsequently, we extend existing tools for estimation of individual ancestry (IA) by allowing for different IA in both parents, leading to estimates of parental individual ancestry (PIA), and a statistical test for recent admixture. Estimation of PIA outperforms IA in most scenarios of recent admixture. Furthermore, additional information about parental ancestry can be acquired with PIA that may guide casework.

LINADMIX: Evaluating the effect of ancient admixture events on modern populations

MOTIVATION

The rise in the number of genotyped ancient individuals provides an opportunity to estimate population admixture models for many populations. However, in models describing modern populations as mixtures of ancient ones, it is typically difficult to estimate the model mixing coefficients and to evaluate its fit to the data.

RESULTS

We present LINADMIX, designed to tackle this problem by solving a constrained linear model when both the ancient and the modern genotypes are represented in a low-dimensional space. LINADMIX estimates the mixing coefficients and their standard errors, and computes a p-value for testing the model fit to the data. We quantified the performance of LINADMIX using an extensive set of simulated studies. We show that LINADMIX can accurately estimate admixture coefficients, and is robust to factors such as population size, genetic drift, proportion of missing data, and various types of model misspecification.

AVAILABILITY

LINADMIX is available as a python code at https://github.com/swidler/linadmix.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.