Quantum processor-inspired machine learning in the biomedical sciences.

Recent advances in high-throughput genomic technologies coupled with exponential increases in computer processing and memory have allowed us to interrogate the complex molecular underpinnings of human disease from a genome-wide perspective. While the deluge of genomic information is expected to increase, a bottleneck in conventional high-performance computing is rapidly approaching. Inspired by recent advances in physical quantum processors, we evaluated several unconventional machine-learning (ML) strategies on actual human tumor data, namely "Ising-type" methods, whose objective function is formulated identical to simulated annealing and quantum annealing. We show the efficacy of multiple Ising-type ML algorithms for classification of multi-omics human cancer data from The Cancer Genome Atlas, comparing these classifiers to a variety of standard ML methods. Our results indicate that Ising-type ML offers superior classification performance with smaller training datasets, thus providing compelling empirical evidence for the potential future application of unconventional computing approaches in the biomedical sciences.

Smooth Muscle Cell Reprogramming in Aortic Aneurysms.

The etiology of aortic aneurysms is poorly understood, but it is associated with atherosclerosis, hypercholesterolemia, and abnormal transforming growth factor ß (TGF-ß) signaling in smooth muscle. Here, we investigated the interactions between these different factors in aortic aneurysm development and identified a key role for smooth muscle cell (SMC) reprogramming into a mesenchymal stem cell (MSC)-like state. SMC-specific ablation of TGF-ß signaling in Apoe-/- mice on a hypercholesterolemic diet led to development of aortic aneurysms exhibiting all the features of human disease, which was associated with transdifferentiation of a subset of contractile SMCs into an MSC-like intermediate state that generated osteoblasts, chondrocytes, adipocytes, and macrophages. This combination of medial SMC loss with marked increases in non-SMC aortic cell mass induced exuberant growth and dilation of the aorta, calcification and ossification of the aortic wall, and inflammation, resulting in aneurysm development.

Chronic mTOR activation induces a degradative smooth muscle cell phenotype.

Smooth muscle cell (SMC) proliferation has been thought to limit the progression of thoracic aortic aneurysm and dissection (TAAD) because loss of medial cells associates with advanced disease. We investigated effects of SMC proliferation in the aortic media by conditional disruption of Tsc1, which hyperactivates mTOR complex 1. Consequent SMC hyperplasia led to progressive medial degeneration and TAAD. In addition to diminished contractile and synthetic functions, fate-mapped SMCs displayed increased proteolysis, endocytosis, phagocytosis, and lysosomal clearance of extracellular matrix and apoptotic cells. SMCs acquired a limited repertoire of macrophage markers and functions via biogenesis of degradative organelles through an mTOR/ß-catenin/MITF-dependent pathway, but were distinguishable from conventional macrophages by an absence of hematopoietic lineage markers and certain immune effectors even in the context of hyperlipidemia. Similar mTOR activation and induction of a degradative SMC phenotype in a model of mild TAAD due to Fbn1 mutation greatly worsened disease with near-uniform lethality. The finding of increased lysosomal markers in medial SMCs from clinical TAAD specimens with hyperplasia and matrix degradation further supports the concept that proliferation of degradative SMCs within the media causes aortic disease, thus identifying mTOR-dependent phenotypic modulation as a therapeutic target for combating TAAD.

Endothelial ERK1/2 signaling maintains integrity of the quiescent endothelium.

To define the role of ERK1/2 signaling in the quiescent endothelium, we induced endothelial Erk2 knockout in adult Erk1-/- mice. This resulted in a rapid onset of hypertension, a decrease in eNOS expression, and an increase in endothelin-1 plasma levels, with all mice dying within 5 wk. Immunostaining and endothelial fate mapping showed a robust increase in TGFß signaling leading to widespread endothelial-to-mesenchymal transition (EndMT). Fibrosis affecting the cardiac conduction system was responsible for the universal lethality in these mice. Other findings included renal endotheliosis, loss of fenestrated endothelia in endocrine organs, and hemorrhages. An ensemble computational intelligence strategy, comprising deep learning and probabilistic programing of RNA-seq data, causally linked the loss of ERK1/2 in HUVECs in vitro to activation of TGFß signaling, EndMT, suppression of eNOS, and induction of endothelin-1 expression. All in silico predictions were verified in vitro and in vivo. In summary, these data establish the key role played by ERK1/2 signaling in the maintenance of vascular normalcy.

Enhancing Retrosynthetic Reaction Prediction with Deep Learning Using Multiscale Reaction Classification.

Chemical synthesis planning is a key aspect in many fields of chemistry, especially drug discovery. Recent implementations of machine learning and artificial intelligence techniques for retrosynthetic analysis have shown great potential to improve computational methods for synthesis planning. Herein, we present a multiscale, data-driven approach for retrosynthetic analysis with deep highway networks (DHN). We automatically extracted reaction rules (i.e., ways in which a molecule is produced) from a data set consisting of chemical reactions derived from U.S. patents. We performed the retrosynthetic reaction prediction task in two steps: first, we built a DHN model to predict which group of reactions (consisting of chemically similar reaction rules) was employed to produce a molecule. Once a reaction group was identified, a DHN trained on the subset of reactions within the identified reaction group, was employed to predict the transformation rule used to produce a molecule. To validate our approach, we predicted the first retrosynthetic reaction step for 40 approved drugs using our multiscale model and compared its predictive performance with a conventional model trained on all machine-extracted reaction rules employed as a control. Our multiscale approach showed a success rate of 82.9% at generating valid reactants from retrosynthetic reaction predictions. Comparatively, the control model trained on all machine-extracted reaction rules yielded a success rate of 58.5% on the validation set of 40 pharmaceutical molecules, indicating a significant statistical improvement with our approach to match known first synthetic reaction of the tested drugs in this study. While our multiscale approach was unable to outperform state-of-the-art rule-based systems curated by expert chemists, multiscale classification represents a marked enhancement in retrosynthetic analysis and can be easily adapted for use in a range of artificial intelligence strategies.

A study based on whole-genome sequencing yields a rare variant at 8q24 associated with prostate cancer.

In Western countries, prostate cancer is the most prevalent cancer of men and one of the leading causes of cancer-related death in men. Several genome-wide association studies have yielded numerous common variants conferring risk of prostate cancer. Here, we analyzed 32.5 million variants discovered by whole-genome sequencing 1,795 Icelanders. We identified a new low-frequency variant at 8q24 associated with prostate cancer in European populations, rs188140481[A] (odds ratio (OR) = 2.90; P(combined) = 6.2 × 10(-34)), with an average risk allele frequency in controls of 0.54%. This variant is only very weakly correlated (r(2) ≤ 0.06) with previously reported risk variants at 8q24, and its association remains significant after adjustment for all known risk-associated variants. Carriers of rs188140481[A] were diagnosed with prostate cancer 1.26 years younger than non-carriers (P = 0.0059). We also report results for a previously described HOXB13 variant (rs138213197[T]), confirming it as a prostate cancer risk variant in populations from across Europe.

Discovery of common variants associated with low TSH levels and thyroid cancer risk.

To search for sequence variants conferring risk of nonmedullary thyroid cancer, we focused our analysis on 22 SNPs with a P < 5 × 10(-8) in a genome-wide association study on levels of thyroid stimulating hormone (TSH) in 27,758 Icelanders. Of those, rs965513 has previously been shown to associate with thyroid cancer. The remaining 21 SNPs were genotyped in 561 Icelandic individuals with thyroid cancer (cases) and up to 40,013 controls. Variants suggestively associated with thyroid cancer (P < 0.05) were genotyped in an additional 595 non-Icelandic cases and 2,604 controls. After combining the results, three variants were shown to associate with thyroid cancer: rs966423 on 2q35 (OR = 1.34; P(combined) = 1.3 × 10(-9)), rs2439302 on 8p12 (OR = 1.36; P(combined) = 2.0 × 10(-9)) and rs116909374 on 14q13.3 (OR = 2.09; P(combined) = 4.6 × 10(-11)), a region previously reported to contain an uncorrelated variant conferring risk of thyroid cancer. A strong association (P = 9.1 × 10(-91)) was observed between rs2439302 on 8p12 and expression of NRG1, which encodes the signaling protein neuregulin 1, in blood.

Genetic correction of PSA values using sequence variants associated with PSA levels.

Measuring serum levels of the prostate-specific antigen (PSA) is the most common screening method for prostate cancer. However, PSA levels are affected by a number of factors apart from neoplasia. Notably, around 40% of the variability of PSA levels in the general population is accounted for by inherited factors, suggesting that it may be possible to improve both sensitivity and specificity by adjusting test results for genetic effects. To search for sequence variants that associate with PSA levels, we performed a genome-wide association study and follow-up analysis using PSA information from 15,757 Icelandic and 454 British men not diagnosed with prostate cancer. Overall, we detected a genome-wide significant association between PSA levels and single-nucleotide polymorphisms (SNPs) at six loci: 5p15.33 (rs2736098), 10q11 (rs10993994), 10q26 (rs10788160), 12q24 (rs11067228), 17q12 (rs4430796), and 19q13.33 [rs17632542 (KLK3: I179T)], each with P(combined) <3 × 10(-10). Among 3834 men who underwent a biopsy of the prostate, the 10q26, 12q24, and 19q13.33 alleles that associate with high PSA levels are associated with higher probability of a negative biopsy (odds ratio between 1.15 and 1.27). Assessment of association between the six loci and prostate cancer risk in 5325 cases and 41,417 controls from Iceland, the Netherlands, Spain, Romania, and the United States showed that the SNPs at 10q26 and 12q24 were exclusively associated with PSA levels, whereas the other four loci also were associated with prostate cancer risk. We propose that a personalized PSA cutoff value, based on genotype, should be used when deciding to perform a prostate biopsy.

The chromosome 9p21 risk locus is associated with angiographic severity and progression of coronary artery disease.

AIMS: we tested the hypothesis that the 9p21 risk locus promotes atherosclerosis by examining the association between rs10757278 and coronary artery disease (CAD) severity and progression determined by semi-quantitative angiographic scores. METHODS AND RESULTS: the rs10757278 single nucleotide polymorphism (SNP) was genotyped as the marker for the 9p21 locus in 2334 Caucasian patients undergoing cardiac catheterization (mean age 63, male 67%). Angiographic CAD was assessed using two semi-quantitative scoring systems with one estimating severity (Gensini) and the other extent (Sullivan). A subset of 308 patients who underwent two or more coronary angiograms at least 6 months apart were examined for net change in Gensini and Sullivan scores over time to determine the rate of CAD progression by genotype and were further classified as 'progressors' or 'non-progressors' based on absolute change per year in angiographic severity score. We replicated the association between the rs10757278 SNP and myocardial infarction and binary (presence/absence) angiographic classifications of CAD. Furthermore, we observed a significant additive association with this SNP, and both severity and extent of CAD using angiographic scores, after adjustment for age, gender, body mass index, traditional cardiovascular risk factors, myocardial infarction, and statin use (Gensini P = 0.016, Sullivan P = 0.005). In addition, there was a significant linear association with CAD progression before and after adjustment for covariates (Gensini P = 0.023, Sullivan P = 0.003) with homozygotes for the risk variant having three-fold greater odds of CAD progression compared with the referent group. CONCLUSION: the 9p21 risk locus is associated with angiographically defined severity, extent, and progression of CAD, suggesting a role for this locus in influencing atherosclerosis and its progression.

Ancestry-shift refinement mapping of the C6orf97-ESR1 breast cancer susceptibility locus.

We used an approach that we term ancestry-shift refinement mapping to investigate an association, originally discovered in a GWAS of a Chinese population, between rs2046210[T] and breast cancer susceptibility. The locus is on 6q25.1 in proximity to the C6orf97 and estrogen receptor alpha (ESR1) genes. We identified a panel of SNPs that are correlated with rs2046210 in Chinese, but not necessarily so in other ancestral populations, and genotyped them in breast cancer case:control samples of Asian, European, and African origin, a total of 10,176 cases and 13,286 controls. We found that rs2046210[T] does not confer substantial risk of breast cancer in Europeans and Africans (OR = 1.04, P = 0.099, and OR = 0.98, P = 0.77, respectively). Rather, in those ancestries, an association signal arises from a group of less common SNPs typified by rs9397435. The rs9397435[G] allele was found to confer risk of breast cancer in European (OR = 1.15, P = 1.2 x 10(-3)), African (OR = 1.35, P = 0.014), and Asian (OR = 1.23, P = 2.9 x 10(-4)) population samples. Combined over all ancestries, the OR was 1.19 (P = 3.9 x 10(-7)), was without significant heterogeneity between ancestries (P(het) = 0.36) and the SNP fully accounted for the association signal in each ancestry. Haplotypes bearing rs9397435[G] are well tagged by rs2046210[T] only in Asians. The rs9397435[G] allele showed associations with both estrogen receptor positive and estrogen receptor negative breast cancer. Using early-draft data from the 1,000 Genomes project, we found that the risk allele of a novel SNP (rs77275268), which is closely correlated with rs9397435, disrupts a partially methylated CpG sequence within a known CTCF binding site. These studies demonstrate that shifting the analysis among ancestral populations can provide valuable resolution in association mapping.

Evaluating differences in linkage disequilibrium between populations.

We propose two methods to evaluate the statistical significance of differences in linkage disequilibrium (LD) between populations, where LD is measured by the standardised parameter D'. The first method is based on bootstrapping individuals within populations in order to test LD differences for each pair of loci. Using this approach we propose a solution to the problem of testing multiple locus-pairs by means of a single test for the number of pairs that exhibit significant LD differences among populations. The second method provides the Bayesian posterior probability that one population has greater LD than the other for each locus pair. Both methods can handle genotypes with unknown phase, and are demonstrated using two data sets. For the purpose of demonstration, we apply the methods to two different sets of data from humans. First, we explore the issue of LD differences between reproductively isolated populations using a new data set of twelve Xq25 microsatellites, typed in four European populations. Second, we examine evidence for LD differences between Alzheimer cases and controls from the Icelandic population using 19 single nucleotide polymorphisms (SNPs) from a 97 kb region flanking the Apolipoprotein E (APOE) gene on chromosome 19.

Addictions and their familiality in Iceland.

Here, we provide an overview of previous family studies of addiction and present a new family study based on clinical data for more than 19,000 individuals who have been treated for addiction in Iceland over the last three decades. Coupled with the extensive Icelandic genealogy information, this population-based sample provides a unique opportunity for family studies. The relative risk (RR) was determined for up to fifth-degree relatives of probands diagnosed with alcohol, cannabis, sedative, and amphetamine dependence. We observe highly significant RR values for all substances ranging from 2.27 for alcohol to 7.3 for amphetamine, for first-degree relatives, and RRs significantly above 1 for distant relations, where the effect of shared environmental factors is minimized. The magnitude of risk in psychostimulant dependence is particularly striking. These findings emphasize the role of genetics in the etiology of addiction and highlight the importance of substance-specific effects.

Genome-wide association and replication studies identify four variants associated with prostate cancer susceptibility.

We report a prostate cancer genome-wide association follow-on study. We discovered four variants associated with susceptibility to prostate cancer in several European populations: rs10934853[A] (OR = 1.12, P = 2.9 x 10(-10)) on 3q21.3; two moderately correlated (r2 = 0.07) variants, rs16902094[G] (OR = 1.21, P = 6.2 x 10(-15)) and rs445114[T] (OR = 1.14, P = 4.7 x 10(-10)), on 8q24.21; and rs8102476[C] (OR = 1.12, P = 1.6 x 10(-11)) on 19q13.2. We also refined a previous association signal on 11q13 with the SNP rs11228565[A] (OR = 1.23, P = 6.7 x 10(-12)). In a multivariate analysis using 22 prostate cancer risk variants typed in the Icelandic population, we estimated that carriers in the top 1.3% of the risk distribution are at a 2.5 times greater risk of developing the disease than members of the general population.

The Scientific Foundation for personal genomics: recommendations from a National Institutes of Health-Centers for Disease Control and Prevention multidisciplinary workshop.

The increasing availability of personal genomic tests has led to discussions about the validity and utility of such tests and the balance of benefits and harms. A multidisciplinary workshop was convened by the National Institutes of Health and the Centers for Disease Control and Prevention to review the scientific foundation for using personal genomics in risk assessment and disease prevention and to develop recommendations for targeted research. The clinical validity and utility of personal genomics is a moving target with rapidly developing discoveries but little translation research to close the gap between discoveries and health impact. Workshop participants made recommendations in five domains: (1) developing and applying scientific standards for assessing personal genomic tests; (2) developing and applying a multidisciplinary research agenda, including observational studies and clinical trials to fill knowledge gaps in clinical validity and utility; (3) enhancing credible knowledge synthesis and information dissemination to clinicians and consumers; (4) linking scientific findings to evidence-based recommendations for use of personal genomics; and (5) assessing how the concept of personal utility can affect health benefits, costs, and risks by developing appropriate metrics for evaluation. To fulfill the promise of personal genomics, a rigorous multidisciplinary research agenda is needed.

Common variants conferring risk of schizophrenia.

Schizophrenia is a complex disorder, caused by both genetic and environmental factors and their interactions. Research on pathogenesis has traditionally focused on neurotransmitter systems in the brain, particularly those involving dopamine. Schizophrenia has been considered a separate disease for over a century, but in the absence of clear biological markers, diagnosis has historically been based on signs and symptoms. A fundamental message emerging from genome-wide association studies of copy number variations (CNVs) associated with the disease is that its genetic basis does not necessarily conform to classical nosological disease boundaries. Certain CNVs confer not only high relative risk of schizophrenia but also of other psychiatric disorders. The structural variations associated with schizophrenia can involve several genes and the phenotypic syndromes, or the 'genomic disorders', have not yet been characterized. Single nucleotide polymorphism (SNP)-based genome-wide association studies with the potential to implicate individual genes in complex diseases may reveal underlying biological pathways. Here we combined SNP data from several large genome-wide scans and followed up the most significant association signals. We found significant association with several markers spanning the major histocompatibility complex (MHC) region on chromosome 6p21.3-22.1, a marker located upstream of the neurogranin gene (NRGN) on 11q24.2 and a marker in intron four of transcription factor 4 (TCF4) on 18q21.2. Our findings implicating the MHC region are consistent with an immune component to schizophrenia risk, whereas the association with NRGN and TCF4 points to perturbation of pathways involved in brain development, memory and cognition.

Common variants on 9q22.33 and 14q13.3 predispose to thyroid cancer in European populations.

In order to search for sequence variants conferring risk of thyroid cancer we conducted a genome-wide association study in 192 and 37,196 Icelandic cases and controls, respectively, followed by a replication study in individuals of European descent. Here we show that two common variants, located on 9q22.33 and 14q13.3, are associated with the disease. Overall, the strongest association signals were observed for rs965513 on 9q22.33 (OR = 1.75; P = 1.7 x 10(-27)) and rs944289 on 14q13.3 (OR = 1.37; P = 2.0 x 10(-9)). The gene nearest to the 9q22.33 locus is FOXE1 (TTF2) and NKX2-1 (TTF1) is among the genes located at the 14q13.3 locus. Both variants contribute to an increased risk of both papillary and follicular thyroid cancer. Approximately 3.7% of individuals are homozygous for both variants, and their estimated risk of thyroid cancer is 5.7-fold greater than that of noncarriers. In a study on a large sample set from the general population, both risk alleles are associated with low concentrations of thyroid stimulating hormone (TSH), and the 9q22.33 allele is associated with low concentration of thyroxin (T(4)) and high concentration of triiodothyronine (T(3)).

New sequence variants associated with bone mineral density.

In an extended genome-wide association study of bone mineral density among 6,865 Icelanders and a follow-up in 8,510 subjects of European descent, we identified four new genome-wide significant loci. These are near the SOST gene at 17q21, the MARK3 gene at 14q32, the SP7 gene at 12q13 and the TNFRSF11A (RANK) gene at 18q21. Furthermore, nonsynonymous SNPs in the C17orf53, LRP4, ADAM19 and IBSP genes were suggestively associated with bone density.

Multiple genetic loci for bone mineral density and fractures.

BACKGROUND: Bone mineral density influences the risk of osteoporosis later in life and is useful in the evaluation of the risk of fracture. We aimed to identify sequence variants associated with bone mineral density and fracture. METHODS: We performed a quantitative trait analysis of data from 5861 Icelandic subjects (the discovery set), testing for an association between 301,019 single-nucleotide polymorphisms (SNPs) and bone mineral density of the hip and lumbar spine. We then tested for an association between 74 SNPs (most of which were implicated in the discovery set) at 32 loci in replication sets of Icelandic, Danish, and Australian subjects (4165, 2269, and 1491 subjects, respectively). RESULTS: Sequence variants in five genomic regions were significantly associated with bone mineral density in the discovery set and were confirmed in the replication sets (combined P values, 1.2x10(-7) to 2.0x10(-21)). Three regions are close to or within genes previously shown to be important to the biologic characteristics of bone: the receptor activator of nuclear factor-kappaB ligand gene (RANKL) (chromosomal location, 13q14), the osteoprotegerin gene (OPG) (8q24), and the estrogen receptor 1 gene (ESR1) (6q25). The two other regions are close to the zinc finger and BTB domain containing 40 gene (ZBTB40) (1p36) and the major histocompatibility complex region (6p21). The 1p36, 8q24, and 6p21 loci were also associated with osteoporotic fractures, as were loci at 18q21, close to the receptor activator of the nuclear factor-kappaB gene (RANK), and loci at 2p16 and 11p11. CONCLUSIONS: We have discovered common sequence variants that are consistently associated with bone mineral density and with low-trauma fractures in three populations of European descent. Although these variants alone are not clinically useful in the prediction of risk to the individual person, they provide insight into the biochemical pathways underlying osteoporosis.

Genetics of gene expression and its effect on disease.

Common human diseases result from the interplay of many genes and environmental factors. Therefore, a more integrative biology approach is needed to unravel the complexity and causes of such diseases. To elucidate the complexity of common human diseases such as obesity, we have analysed the expression of 23,720 transcripts in large population-based blood and adipose tissue cohorts comprehensively assessed for various phenotypes, including traits related to clinical obesity. In contrast to the blood expression profiles, we observed a marked correlation between gene expression in adipose tissue and obesity-related traits. Genome-wide linkage and association mapping revealed a highly significant genetic component to gene expression traits, including a strong genetic effect of proximal (cis) signals, with 50% of the cis signals overlapping between the two tissues profiled. Here we demonstrate an extensive transcriptional network constructed from the human adipose data that exhibits significant overlap with similar network modules constructed from mouse adipose data. A core network module in humans and mice was identified that is enriched for genes involved in the inflammatory and immune response and has been found to be causally associated to obesity-related traits.