Artificial Intelligence in Hypertension: Seeing Through a Glass Darkly.

Hypertension remains the largest modifiable cause of mortality worldwide despite the availability of effective medications and sustained research efforts over the past 100 years. Hypertension requires transformative solutions that can help reduce the global burden of the disease. Artificial intelligence and machine learning, which have made a substantial impact on our everyday lives over the last decade may be the route to this transformation. However, artificial intelligence in health care is still in its nascent stages and realizing its potential requires numerous challenges to be overcome. In this review, we provide a clinician-centric perspective on artificial intelligence and machine learning as applied to medicine and hypertension. We focus on the main roadblocks impeding implementation of this technology in clinical care and describe efforts driving potential solutions. At the juncture, there is a critical requirement for clinical and scientific expertise to work in tandem with algorithmic innovation followed by rigorous validation and scrutiny to realize the promise of artificial intelligence-enabled health care for hypertension and other chronic diseases.

Cluster analysis of angiotensin biomarkers to identify antihypertensive drug treatment in population studies.

BACKGROUND: The recent progress in molecular biology generates an increasing interest in investigating molecular biomarkers as markers of response to treatments. The present work is motivated by a study, where the objective was to explore the potential of the molecular biomarkers of renin-angiotensin-aldosterone system (RAAS) to identify the undertaken antihypertensive treatments in the general population. Population-based studies offer an opportunity to assess the effectiveness of treatments in real-world scenarios. However, lack of quality documentation, especially when electronic health record linkage is unavailable, leads to inaccurate reporting and classification bias. METHOD: We present a machine learning clustering technique to determine the potential of measured RAAS biomarkers for the identification of undertaken treatments in the general population. The biomarkers were simultaneously determined through a novel mass-spectrometry analysis in 800 participants of the Cooperative Health Research In South Tyrol (CHRIS) study with documented antihypertensive treatments. We assessed the agreement, sensitivity and specificity of the resulting clusters against known treatment types. Through the lasso penalized regression, we identified clinical characteristics associated with the biomarkers, accounting for the effects of cluster and treatment classifications. RESULTS: We identified three well-separated clusters: cluster 1 (n = 444) preferentially including individuals not receiving RAAS-targeting drugs; cluster 2 (n = 235) identifying angiotensin type 1 receptor blockers (ARB) users (weighted kappa κw = 74%; sensitivity = 73%; specificity = 83%); and cluster 3 (n = 121) well discriminating angiotensin-converting enzyme inhibitors (ACEi) users (κw = 81%; sensitivity = 55%; specificity = 90%). Individuals in clusters 2 and 3 had higher frequency of diabetes as well as higher fasting glucose and BMI levels. Age, sex and kidney function were strong predictors of the RAAS biomarkers independently of the cluster structure. CONCLUSIONS: Unsupervised clustering of angiotensin-based biomarkers is a viable technique to identify individuals on specific antihypertensive treatments, pointing to a potential application of the biomarkers as useful clinical diagnostic tools even outside of a controlled clinical setting.

Meta-analysis of up to 622,409 individuals identifies 40 novel smoking behaviour associated genetic loci.

Smoking is a major heritable and modifiable risk factor for many diseases, including cancer, common respiratory disorders and cardiovascular diseases. Fourteen genetic loci have previously been associated with smoking behaviour-related traits. We tested up to 235,116 single nucleotide variants (SNVs) on the exome-array for association with smoking initiation, cigarettes per day, pack-years, and smoking cessation in a fixed effects meta-analysis of up to 61 studies (up to 346,813 participants). In a subset of 112,811 participants, a further one million SNVs were also genotyped and tested for association with the four smoking behaviour traits. SNV-trait associations with P < 5 × 10-8 in either analysis were taken forward for replication in up to 275,596 independent participants from UK Biobank. Lastly, a meta-analysis of the discovery and replication studies was performed. Sixteen SNVs were associated with at least one of the smoking behaviour traits (P < 5 × 10-8) in the discovery samples. Ten novel SNVs, including rs12616219 near TMEM182, were followed-up and five of them (rs462779 in REV3L, rs12780116 in CNNM2, rs1190736 in GPR101, rs11539157 in PJA1, and rs12616219 near TMEM182) replicated at a Bonferroni significance threshold (P < 4.5 × 10-3) with consistent direction of effect. A further 35 SNVs were associated with smoking behaviour traits in the discovery plus replication meta-analysis (up to 622,409 participants) including a rare SNV, rs150493199, in CCDC141 and two low-frequency SNVs in CEP350 and HDGFRP2. Functional follow-up implied that decreased expression of REV3L may lower the probability of smoking initiation. The novel loci will facilitate understanding the genetic aetiology of smoking behaviour and may lead to the identification of potential drug targets for smoking prevention and/or cessation.

Discovery of novel heart rate-associated loci using the Exome Chip.

Resting heart rate is a heritable trait, and an increase in heart rate is associated with increased mortality risk. Genome-wide association study analyses have found loci associated with resting heart rate, at the time of our study these loci explained 0.9% of the variation. This study aims to discover new genetic loci associated with heart rate from Exome Chip meta-analyses.Heart rate was measured from either elecrtrocardiograms or pulse recordings. We meta-analysed heart rate association results from 104 452 European-ancestry individuals from 30 cohorts, genotyped using the Exome Chip. Twenty-four variants were selected for follow-up in an independent dataset (UK Biobank, N = 134 251). Conditional and gene-based testing was undertaken, and variants were investigated with bioinformatics methods.We discovered five novel heart rate loci, and one new independent low-frequency non-synonymous variant in an established heart rate locus (KIAA1755). Lead variants in four of the novel loci are non-synonymous variants in the genes C10orf71, DALDR3, TESK2 and SEC31B. The variant at SEC31B is significantly associated with SEC31B expression in heart and tibial nerve tissue. Further candidate genes were detected from long-range regulatory chromatin interactions in heart tissue (SCD, SLF2 and MAPK8). We observed significant enrichment in DNase I hypersensitive sites in fetal heart and lung. Moreover, enrichment was seen for the first time in human neuronal progenitor cells (derived from embryonic stem cells) and fetal muscle samples by including our novel variants.Our findings advance the knowledge of the genetic architecture of heart rate, and indicate new candidate genes for follow-up functional studies.

Salt stress in the renal tubules is linked to TAL-specific expression of uromodulin and an upregulation of heat shock genes.

Previously, our comprehensive cardiovascular characterization study validated Uromodulin as a blood pressure gene. Uromodulin is a glycoprotein exclusively synthesized at the thick ascending limb of the loop of Henle and is encoded by the Umod gene. Umod-/- mice have significantly lower blood pressure than Umod+/+ mice, are resistant to salt-induced changes in blood pressure, and show a leftward shift in pressure-natriuresis curves reflecting changes of sodium reabsorption. Salt stress triggers transcription factors and genes that alter renal sodium reabsorption. To date there are no studies on renal transcriptome responses to salt stress. Here we aimed use RNA-Seq to delineate salt stress pathways in tubules isolated from Umod+/+ mice (a model of sodium retention) and Umod-/- mice (a model of sodium depletion) ± 300 mosmol sodium chloride ( n = 3 per group). In response to salt stress, the tubules of Umod+/+ mice displayed an upregulation of heat shock transcripts. The greatest changes occurred in the expression of: Hspa1a (Log2 fold change 4.35, P = 2.48 e-12) and Hspa1b (Log2 fold change 4.05, P = 2.48 e-12). This response was absent in tubules of Umod-/- mice. Interestingly, seven of the genes discordantly expressed in the Umod-/- tubules were electrolyte transporters. Our results are the first to show that salt stress in renal tubules alters the transcriptome, increasing the expression of heat shock genes. This direction of effect in Umod+/+ tubules suggest the difference is due to the presence of Umod facilitating greater sodium entry into the tubule cell reflecting a specific response to salt stress.

Effects of dietary salt on gene and protein expression in brain tissue of a model of sporadic small vessel disease.

BACKGROUND: The effect of salt on cerebral small vessel disease (SVD) is poorly understood. We assessed the effect of dietary salt on cerebral tissue of the stroke-prone spontaneously hypertensive rat (SHRSP) - a relevant model of sporadic SVD - at both the gene and protein level. Methods: Brains from 21-week-old SHRSP and Wistar-Kyoto rats, half additionally salt-loaded (via a 3-week regime of 1% NaCl in drinking water), were split into two hemispheres and sectioned coronally - one hemisphere for mRNA microarray and qRT-PCR, the other for immunohistochemistry using a panel of antibodies targeting components of the neurovascular unit. Results: We observed differences in gene and protein expression affecting the acute phase pathway and oxidative stress (ALB, AMBP, APOH, AHSG and LOC100129193, up-regulated in salt-loaded WKY versus WKY, >2-fold), active microglia (increased Iba-1 protein expression in salt-loaded SHRSP versus salt-loaded WKY, p<0.05), vascular structure (ACTB and CTNNB, up-regulated in salt-loaded SHRSP versus SHRSP, >3-fold; CLDN-11, VEGF and VGF down-regulated >2-fold in salt-loaded SHRSP versus SHRSP) and myelin integrity (MBP down-regulated in salt loaded WKY rats versus WKY, >2.5-fold). Changes of salt-loading were more pronounced in SHRSP and occurred without an increase in blood pressure in WKY rats. CONCLUSION: Salt exposure induced changes in gene and protein expression in an experimental model of SVD and its parent rat strain in multiple pathways involving components of the glio-vascular unit. Further studies in pertinent experimental models at different ages would help clarify the short- and long-term effect of dietary salt in SVD.

Recent Findings in the Genetics of Blood Pressure: How to Apply in Practice or Is a Moonshot Required?

PURPOSE OF REVIEW: Hypertension is recognised as the biggest contributor to the global burden of disease, but it is controlled in less than a fifth of patients worldwide, despite being relatively easy to detect and the availability of inexpensive safe generic drugs. Blood pressure is regulated by a complex network of physiologic pathways with currently available drugs targeting key receptors or enzymes in the top pathways. Major advances in the dissection of both monogenic and polygenic determinants of blood pressure regulation and variation have not resulted in rapid translation of these discoveries into clinical applications or precision medicine. RECENT FINDINGS: Uromodulin is an example of a novel gene for hypertension identified from genome-wide association studies, currently the basis of a clinical trial to reposition loop diuretics in hypertension management. Gene-editing studies have established a genome-wide association studies (GWAS) SNP in chromosome 6p24, implicated in six conditions including hypertension, as a distal regulator of the endothelin-1 gene around 3000 base pairs away. Genomics of aldosterone-producing adenomas bring to focus the paradox in genomic medicine where availability of cheap generic drugs may render precision medicine uneconomical. The speed of technology-driven genomic discoveries and the sluggish traditional pathways of drug development and translation need harmonisation to make a timely and early impact on global public health. This requires a directed collaborative effort for which we propose a hypertension moonshot to make a quantum leap in hypertension management and cardiovascular risk reduction by bringing together traditional bioscience, omics, engineering, digital technology and data science.

Role of renal transporters and novel regulatory interactions in the TAL that control blood pressure.

Hypertension (HTN), a major public health issue is currently the leading factor in the global burden of disease, where associated complications account for 9.4 million deaths worldwide every year. Excessive dietary salt intake is among the environmental factors that contribute to HTN, known as salt sensitivity. The heterogeneity of salt sensitivity and the multiple mechanisms that link high salt intake to increases in blood pressure are of upmost importance for therapeutic application. A continual increase in the kidney's reabsorption of sodium (Na+) relies on sequential actions at various segments along the nephron. When the distal segments of the nephron fail to regulate Na+, the effects on Na+ homeostasis are unfavorable. We propose that the specific nephron region where increased active uptake occurs as a result of variations in Na+ reabsorption is at the thick ascending limb of the loop of Henle (TAL). The purpose of this review is to urge the consideration of the TAL as contributing to the pathophysiology of salt-sensitive HTN. Further research in this area will enable development of a therapeutic application for targeted treatment.

Mosaic structural variation in children with developmental disorders.

Delineating the genetic causes of developmental disorders is an area of active investigation. Mosaic structural abnormalities, defined as copy number or loss of heterozygosity events that are large and present in only a subset of cells, have been detected in 0.2-1.0% of children ascertained for clinical genetic testing. However, the frequency among healthy children in the community is not well characterized, which, if known, could inform better interpretation of the pathogenic burden of this mutational category in children with developmental disorders. In a case-control analysis, we compared the rate of large-scale mosaicism between 1303 children with developmental disorders and 5094 children lacking developmental disorders, using an analytical pipeline we developed, and identified a substantial enrichment in cases (odds ratio = 39.4, P-value 1.073e - 6). A meta-analysis that included frequency estimates among an additional 7000 children with congenital diseases yielded an even stronger statistical enrichment (P-value 1.784e - 11). In addition, to maximize the detection of low-clonality events in probands, we applied a trio-based mosaic detection algorithm, which detected two additional events in probands, including an individual with genome-wide suspected chimerism. In total, we detected 12 structural mosaic abnormalities among 1303 children (0.9%). Given the burden of mosaicism detected in cases, we suspected that many of the events detected in probands were pathogenic. Scrutiny of the genotypic-phenotypic relationship of each detected variant assessed that the majority of events are very likely pathogenic. This work quantifies the burden of structural mosaicism as a cause of developmental disorders.

Genetic and molecular aspects of hypertension.

Until recently, significant advances in our understanding of the mechanisms of blood pressure regulation arose from studies of monogenic forms of hypertension and hypotension, which identified rare variants that primarily alter renal salt handling. Genome-wide association and exome sequencing studies over the past 6 years have resulted in an unparalleled burst of discovery in the genetics of blood pressure regulation and hypertension. More importantly, genome-wide association studies, while expanding the list of common genetic variants associated with blood pressure and hypertension, are also uncovering novel pathways of blood pressure regulation that augur a new era of novel drug development, repurposing, and stratification in the management of hypertension. In this review, we describe the current state of the art of the genetic and molecular basis of blood pressure and hypertension.

Genomics of hypertension.

A complex network of interacting pathways involving renal, neural, endocrine, vascular and other mechanisms controls the main determinants of blood pressure - cardiac output and total peripheral resistance. Multiple genes within each of these systems contribute to the specialized functions regulating blood pressure. The monogenic forms of blood pressure dysregulation have provided valuable insights into blood pressure regulation and expanded our understanding of both the mechanisms and the treatment of hypertension. Genome wide association studies have identified over 100 single nucleotide polymorphisms associated with blood pressure phenotypes and have identified plausible novel pathways of BP regulation and putative drug targets.

Contrasting mortality risks among subgroups of treated hypertensive patients developing new-onset diabetes.

AIMS: Hypertension and diabetes mellitus (DM) frequently cluster together and synergistically increase cardiovascular risk. Among those who develop DM during treatment for hypertension (new-onset diabetes, NOD), it is unclear whether NOD reflects a separate entity associated with increased risk or merely reflects accelerated presentation of DM. METHODS AND RESULTS: We analysed data on 15 089 hypertensive patients attending the Glasgow Blood Pressure Clinic. The date at first hospital encounter either with diagnosis of diabetes or prescription of anti-hyperglycaemic medication were considered as the onset of diabetes. Cox proportional hazard models (including propensity score matching) were employed to study associations between diabetes status, early and late NOD (diagnosis <10 years or >10 years from first clinic visit) and cause-specific mortality. There were 2516 patients (16.7%) with DM, of whom 1862 (12.3%) had NOD [early NOD = 705 (4.6%); late NOD = 1157 (7.6%)]. The incidence rate of NOD was 8.2 per 1000 person-years. The total time at risk was 239 929 person-years [median survival: 28.1 years (inter-quartile range: 16.2-39.9)]. Compared with non-diabetic individuals, prevalent DM [hazard ratio (HR) = 1.8, 95% confidence interval (CI): 1.4-2.2] and time varying NOD status (HR: 1.09, 95% CI: 1.06-1.17) were associated with increased adjusted all-cause mortality. Early NOD (HR: 1.39, 95% CI: 1.2-1.6) was associated with increased in mortality risk, but not late NOD (HR: 0.92, 95% CI: 0.83-1.01). Results were consistent in the propensity score matched analyses. CONCLUSION: Although 1-in-8 hypertensive patients develop NOD, mortality is increased only in the 1-in-20 who develop early NOD. Further studies are warranted to determine if early identification of such individuals should provide an alert for intensification of therapeutic interventions.

Urine proteomics in the diagnosis of stable angina.

BACKGROUND: We have previously described a panel of 238 urinary polypeptides specific for established severe coronary artery disease (CAD). Here we studied this polypeptide panel in patients with a wider range of CAD severity. METHODS: We recruited 60 patients who underwent elective coronary angiography for investigation of stable angina. Patients were selected for either having angiographic evidence of CAD or not (NCA) following coronary angiography (n = 30/30; age, 55 ± 6 vs. 56 ± 7 years, P = 0.539) to cover the extremes of the CAD spectrum. A further 66 patients with severe CAD (age, 64 ± 9 years) prior to surgical coronary revascularization were added for correlation studies. The Gensini score was calculated from coronary angiograms as a measure of CAD severity. Urinary proteomic analyses were performed using capillary electrophoresis coupled online to micro time-of-flight mass spectrometry. The urinary polypeptide pattern was classified using a predefined algorithm and resulting in the CAD238 score, which expresses the pattern quantitatively. RESULTS: In the whole cohort of patients with CAD (Gensini score 60 [40; 98]) we found a close correlation between Gensini scores and CAD238 (ρ = 0.465, P < 0.001). After adjustment for age (ß = 0.144; P = 0.135) the CAD238 score remained a significant predictor of the Gensini score (ß =0.418; P < 0.001). In those with less severe CAD (Gensini score 40 [25; 61]), however, we could not detect a difference in CAD238 compared to patients with NCA (-0.487 ± 0.341 vs. -0.612 ± 0.269, P = 0.119). CONCLUSIONS: In conclusion the urinary polypeptide CAD238 score is associated with CAD burden and has potential as a new cardiovascular biomarker.

Genome-wide association study of CNVs in 16,000 cases of eight common diseases and 3,000 shared controls.

Copy number variants (CNVs) account for a major proportion of human genetic polymorphism and have been predicted to have an important role in genetic susceptibility to common disease. To address this we undertook a large, direct genome-wide study of association between CNVs and eight common human diseases. Using a purpose-designed array we typed approximately 19,000 individuals into distinct copy-number classes at 3,432 polymorphic CNVs, including an estimated approximately 50% of all common CNVs larger than 500 base pairs. We identified several biological artefacts that lead to false-positive associations, including systematic CNV differences between DNAs derived from blood and cell lines. Association testing and follow-up replication analyses confirmed three loci where CNVs were associated with disease-IRGM for Crohn's disease, HLA for Crohn's disease, rheumatoid arthritis and type 1 diabetes, and TSPAN8 for type 2 diabetes-although in each case the locus had previously been identified in single nucleotide polymorphism (SNP)-based studies, reflecting our observation that most common CNVs that are well-typed on our array are well tagged by SNPs and so have been indirectly explored through SNP studies. We conclude that common CNVs that can be typed on existing platforms are unlikely to contribute greatly to the genetic basis of common human diseases.

Renal Mechanisms of Association between Fibroblast Growth Factor 1 and Blood Pressure.

The fibroblast growth factor 1 (FGF1) gene is expressed primarily in the kidney and may contribute to hypertension. However, the biologic mechanisms underlying the association between FGF1 and BP regulation remain unknown. We report that the major allele of FGF1 single nucleotide polymorphism rs152524 was associated in a dose-dependent manner with systolic BP (P = 9.65 × 10(-5)) and diastolic BP (P = 7.61 × 10(-3)) in a meta-analysis of 14,364 individuals and with renal expression of FGF1 mRNA in 126 human kidneys (P=9.0 × 10(-3)). Next-generation RNA sequencing revealed that upregulated renal expression of FGF1 or of each of the three FGF1 mRNA isoforms individually was associated with higher BP. FGF1-stratified coexpression analysis in two separate collections of human kidneys identified 126 FGF1 partner mRNAs, of which 71 and 63 showed at least nominal association with systolic and diastolic BP, respectively. Of those mRNAs, seven mRNAs in five genes (MME, PTPRO, REN, SLC12A3, and WNK1) had strong prior annotation to BP or hypertension. MME, which encodes an enzyme that degrades circulating natriuretic peptides, showed the strongest differential coexpression with FGF1 between hypertensive and normotensive kidneys. Furthermore, higher level of renal FGF1 expression was associated with lower circulating levels of atrial and brain natriuretic peptides. These findings indicate that FGF1 expression in the kidney is at least under partial genetic control and that renal expression of several FGF1 partner genes involved in the natriuretic peptide catabolism pathway, renin-angiotensin cascade, and sodium handling network may explain the association between FGF1 and BP.

Mendelian randomization of blood lipids for coronary heart disease.

AIMS: To investigate the causal role of high-density lipoprotein cholesterol (HDL-C) and triglycerides in coronary heart disease (CHD) using multiple instrumental variables for Mendelian randomization. METHODS AND RESULTS: We developed weighted allele scores based on single nucleotide polymorphisms (SNPs) with established associations with HDL-C, triglycerides, and low-density lipoprotein cholesterol (LDL-C). For each trait, we constructed two scores. The first was unrestricted, including all independent SNPs associated with the lipid trait identified from a prior meta-analysis (threshold P < 2 × 10(-6)); and the second a restricted score, filtered to remove any SNPs also associated with either of the other two lipid traits at P ≤ 0.01. Mendelian randomization meta-analyses were conducted in 17 studies including 62,199 participants and 12,099 CHD events. Both the unrestricted and restricted allele scores for LDL-C (42 and 19 SNPs, respectively) associated with CHD. For HDL-C, the unrestricted allele score (48 SNPs) was associated with CHD (OR: 0.53; 95% CI: 0.40, 0.70), per 1 mmol/L higher HDL-C, but neither the restricted allele score (19 SNPs; OR: 0.91; 95% CI: 0.42, 1.98) nor the unrestricted HDL-C allele score adjusted for triglycerides, LDL-C, or statin use (OR: 0.81; 95% CI: 0.44, 1.46) showed a robust association. For triglycerides, the unrestricted allele score (67 SNPs) and the restricted allele score (27 SNPs) were both associated with CHD (OR: 1.62; 95% CI: 1.24, 2.11 and 1.61; 95% CI: 1.00, 2.59, respectively) per 1-log unit increment. However, the unrestricted triglyceride score adjusted for HDL-C, LDL-C, and statin use gave an OR for CHD of 1.01 (95% CI: 0.59, 1.75). CONCLUSION: The genetic findings support a causal effect of triglycerides on CHD risk, but a causal role for HDL-C, though possible, remains less certain.

Mutations in the gene encoding the 3'-5' DNA exonuclease TREX1 are associated with systemic lupus erythematosus.

TREX1 acts in concert with the SET complex in granzyme A-mediated apoptosis, and mutations in TREX1 cause Aicardi-Goutières syndrome and familial chilblain lupus. Here, we report monoallelic frameshift or missense mutations and one 3' UTR variant of TREX1 present in 9/417 individuals with systemic lupus erythematosus but absent in 1,712 controls (P = 4.1 x 10(-7)). We demonstrate that two mutant TREX1 alleles alter subcellular targeting. Our findings implicate TREX1 in the pathogenesis of SLE.

Genetic basis of blood pressure and hypertension.

Blood pressure (BP) is a complex trait regulated by an intricate network of physiological pathways involving extracellular fluid volume homeostasis, cardiac contractility and vascular tone through renal, neural or endocrine systems. Untreated high BP, or hypertension (HTN), is associated with increased mortality, and thus a better understanding of the pathophysiological and genetic underpinnings of BP regulation will have a major impact on public health. However, identifying genes that contribute to BP and HTN has proved challenging. In this review we describe our current understanding of the genetic architecture of BP and HTN, which has accelerated over the past five years primarily owing to genome-wide association studies (GWAS) and the continuing progress in uncovering rare gene mutations, epigenetic markers and regulatory pathways involved in the physiology of BP. We also look ahead to future studies characterizing novel pathways that affect BP and HTN and discuss strategies for translating current findings to the clinic.

Copy number variation in the human Y chromosome in the UK population.

We have assessed copy number variation (CNV) in the male-specific part of the human Y chromosome discovered by array comparative genomic hybridization (array-CGH) in 411 apparently healthy UK males, and validated the findings using SNP genotype intensity data available for 149 of them. After manual curation taking account of the complex duplicated structure of Y-chromosomal sequences, we discovered 22 curated CNV events considered validated or likely, mean 0.93 (range 0-4) per individual. 16 of these were novel. Curated CNV events ranged in size from <1 kb to >3 Mb, and in frequency from 1/411 to 107/411. Of the 24 protein-coding genes or gene families tested, nine showed CNV. These included a large duplication encompassing the AMELY and TBL1Y genes that probably has no phenotypic effect, partial deletions of the TSPY cluster and AZFc region that may influence spermatogenesis, and other variants with unknown functional implications, including abundant variation in the number of RBMY genes and/or pseudogenes, and a novel complex duplication of two segments overlapping the AZFa region and including the 3' end of the UTY gene.