Artificial intelligence enables comprehensive genome interpretation and nomination of candidate diagnoses for rare genetic diseases.

BACKGROUND: Clinical interpretation of genetic variants in the context of the patient's phenotype is becoming the largest component of cost and time expenditure for genome-based diagnosis of rare genetic diseases. Artificial intelligence (AI) holds promise to greatly simplify and speed genome interpretation by integrating predictive methods with the growing knowledge of genetic disease. Here we assess the diagnostic performance of Fabric GEM, a new, AI-based, clinical decision support tool for expediting genome interpretation. METHODS: We benchmarked GEM in a retrospective cohort of 119 probands, mostly NICU infants, diagnosed with rare genetic diseases, who received whole-genome or whole-exome sequencing (WGS, WES). We replicated our analyses in a separate cohort of 60 cases collected from five academic medical centers. For comparison, we also analyzed these cases with current state-of-the-art variant prioritization tools. Included in the comparisons were trio, duo, and singleton cases. Variants underpinning diagnoses spanned diverse modes of inheritance and types, including structural variants (SVs). Patient phenotypes were extracted from clinical notes by two means: manually and using an automated clinical natural language processing (CNLP) tool. Finally, 14 previously unsolved cases were reanalyzed. RESULTS: GEM ranked over 90% of the causal genes among the top or second candidate and prioritized for review a median of 3 candidate genes per case, using either manually curated or CNLP-derived phenotype descriptions. Ranking of trios and duos was unchanged when analyzed as singletons. In 17 of 20 cases with diagnostic SVs, GEM identified the causal SVs as the top candidate and in 19/20 within the top five, irrespective of whether SV calls were provided or inferred ab initio by GEM using its own internal SV detection algorithm. GEM showed similar performance in absence of parental genotypes. Analysis of 14 previously unsolved cases resulted in a novel finding for one case, candidates ultimately not advanced upon manual review for 3 cases, and no new findings for 10 cases. CONCLUSIONS: GEM enabled diagnostic interpretation inclusive of all variant types through automated nomination of a very short list of candidate genes and disorders for final review and reporting. In combination with deep phenotyping by CNLP, GEM enables substantial automation of genetic disease diagnosis, potentially decreasing cost and expediting case review.

Toward clinical genomics in everyday medicine: perspectives and recommendations.

Precision or personalized medicine through clinical genome and exome sequencing has been described by some as a revolution that could transform healthcare delivery, yet it is currently used in only a small fraction of patients, principally for the diagnosis of suspected Mendelian conditions and for targeting cancer treatments. Given the burden of illness in our society, it is of interest to ask how clinical genome and exome sequencing can be constructively integrated more broadly into the routine practice of medicine for the betterment of public health. In November 2014, 46 experts from academia, industry, policy and patient advocacy gathered in a conference sponsored by Illumina, Inc. to discuss this question, share viewpoints and propose recommendations. This perspective summarizes that work and identifies some of the obstacles and opportunities that must be considered in translating advances in genomics more widely into the practice of medicine.

Clinical application of pharmacogenetics: focusing on practical issues.

Recent large-scale genetic-based studies have transformed the field of pharmacogenetics to identify, characterize and leverage genetic information to inform patient care. Genetic testing can be used to alter drug selection, optimize drug dosing and prevent unnecessary adverse events. As precision medicine becomes the mainstay in the clinic, it becomes critical for clinicians to utilize pharmacogenetics to guide patient care. One primary challenge is identifying patients where genetic tests that can potentially impact patient care. To address this challenge, our review highlights many practical issues clinicians may encounter: identifying candidate patients and clinical laboratories for pharmacogenetic testing, selecting highly curated resources to help asses test validity, reimbursing costs of pharmacogenetic tests, and interpreting of pharmacogenetic test results.

Genomic medicine: a decade of successes, challenges, and opportunities.

Genomic medicine--an aspirational term 10 years ago--is gaining momentum across the entire clinical continuum from risk assessment in healthy individuals to genome-guided treatment in patients with complex diseases. We review the latest achievements in genome research and their impact on medicine, primarily in the past decade. In most cases, genomic medicine tools remain in the realm of research, but some tools are crossing over into clinical application, where they have the potential to markedly alter the clinical care of patients. In this State of the Art Review, we highlight notable examples including the use of next-generation sequencing in cancer pharmacogenomics, in the diagnosis of rare disorders, and in the tracking of infectious disease outbreaks. We also discuss progress in dissecting the molecular basis of common diseases, the role of the host microbiome, the identification of drug response biomarkers, and the repurposing of drugs. The significant challenges of implementing genomic medicine are examined, along with the innovative solutions being sought. These challenges include the difficulty in establishing clinical validity and utility of tests, how to increase awareness and promote their uptake by clinicians, a changing regulatory and coverage landscape, the need for education, and addressing the ethical aspects of genomics for patients and society. Finally, we consider the future of genomics in medicine and offer a glimpse of the forces shaping genomic medicine, such as fundamental shifts in how we define disease, how medicine is delivered to patients, and how consumers are managing their own health and affecting change.

Development and evaluation of a genetic risk score for obesity.

Multi-locus profiles of genetic risk, so-called "genetic risk scores," can be used to translate discoveries from genome-wide association studies into tools for population health research. We developed a genetic risk score for obesity from results of 16 published genome-wide association studies of obesity phenotypes in European-descent samples. We then evaluated this genetic risk score using data from the Atherosclerosis Risk in Communities (ARIC) cohort GWAS sample (N = 10,745, 55% female, 77% white, 23% African American). Our 32-locus GRS was a statistically significant predictor of body mass index (BMI) and obesity among ARIC whites [for BMI, r = 0.13, p<1 × 10(-30); for obesity, area under the receiver operating characteristic curve (AUC) = 0.57 (95% CI 0.55-0.58)]. The GRS predicted differences in obesity risk net of demographic, geographic, and socioeconomic information. The GRS performed less well among African Americans. The genetic risk score we derived from GWAS provides a molecular measurement of genetic predisposition to elevated BMI and obesity.[Supplemental materials are available for this article. Go to the publisher's online edition of Biodemography and Social Biology for the following resource: Supplement to Development & Evaluation of a Genetic Risk Score for Obesity.].

Pleiotropy and allelic heterogeneity in the TOMM40-APOE genomic region related to clinical and metabolic features of hepatitis C infection.

Hepatitis C virus (HCV) modulates host lipid metabolism as part of its lifecycle and is dependent upon VLDL for co-assembly and secretion. HCV dyslipidemia is associated with steatosis, insulin resistance, IL28B genotype and disease progression. Apolipoprotein E (ApoE) is an important lipid transport protein, a key constituent of VLDL, and is involved in immunomodulation. Our aims were to determine the role of APOE regional polymorphisms on host lipids, IL28B genotype and disease severity in chronic HCV (CHC) patients. The study cohort included 732 CHC patients with available DNA for genotype determination of four polymorphisms in the chromosome 19 region that encompasses the TOMM40, APOE and APOC1 genes. Serum lipid analysis and apolipoproteins levels were measured using an immunoturbidimetric assay. APOE rs7412 polymorphism (capturing the ε2 isoform) was significantly associated with serum ApoE levels in both Caucasians and African-American patients (p = 2.3 × 10(-11)) and explained 7 % of variance in serum ApoE. Among IL28B-CC patients (n = 196), the rs429358 (defines ε4 isoform) and TOMM40 '523' S polymorphisms were associated with 12 % of variance in ApoB levels. Patients homozygous for the APOE ε3 isoform had a greater than twofold increased odds of F2-F4 fibrosis (p = 1.8 × 10(-5)), independent of serum lipid and lipoprotein levels. There were no associations between APOE polymorphisms and serum HDL-C, APO-CIII and triglycerides. In CHC patients, genetic heterogeneity in the APOE/TOMM40 genomic region is significantly associated with variation in serum ApoE and ApoB levels, and also with fibrosis suggesting a pleiotropic attribute of this genomic region.

Metaprotein expression modeling for label-free quantitative proteomics.

BACKGROUND: Label-free quantitative proteomics holds a great deal of promise for the future study of both medicine and biology. However, the data generated is extremely intricate in its correlation structure, and its proper analysis is complex. There are issues with missing identifications. There are high levels of correlation between many, but not all, of the peptides derived from the same protein. Additionally, there may be systematic shifts in the sensitivity of the machine between experiments or even through time within the duration of a single experiment. RESULTS: We describe a hierarchical model for analyzing unbiased, label-free proteomics data which utilizes the covariance of peptide expression across samples as well as MS/MS-based identifications to group peptides-a strategy we call metaprotein expression modeling. Our metaprotein model acknowledges the possibility of misidentifications, post-translational modifications and systematic differences between samples due to changes in instrument sensitivity or differences in total protein concentration. In addition, our approach allows us to validate findings from unbiased, label-free proteomics experiments with further unbiased, label-free proteomics experiments. Finally, we demonstrate the clinical/translational utility of the model for building predictors capable of differentiating biological phenotypes as well as for validating those findings in the context of three novel cohorts of patients with Hepatitis C. CONCLUSIONS: Mass-spectrometry proteomics is quickly becoming a powerful tool for studying biological and translational questions. Making use of all of the information contained in a particular set of data will be critical to the success of those endeavors. Our proposed model represents an advance in the ability of statistical models of proteomic data to identify and utilize correlation between features. This allows validation of predictors without translation to targeted assays in addition to informing the choice of targets when it is appropriate to generate those assays.

Characterization of the poly-T variant in the TOMM40 gene in diverse populations.

We previously discovered that a polymorphic, deoxythymidine-homopolymer (poly-T, rs10524523) in intron 6 of the TOMM40 gene is associated with age-of-onset of Alzheimer's disease and with cognitive performance in elderly. Three allele groups were defined for rs10524523, hereafter '523', based on the number of 'T'-residues: 'Short' (S, T≤19), 'Long' (L, 20≤T≤29) and 'Very Long' (VL, T≥30). Homopolymers, particularly long homopolymers like '523', are difficult to genotype because 'slippage' occurs during PCR-amplification. We initially genotyped this locus by PCR-amplification followed by Sanger-sequencing. However, we recognized the need to develop a higher-throughput genotyping method that is also accurate and reliable. Here we describe a new '523' genotyping assay that is simple and inexpensive to perform in a standard molecular genetics laboratory. The assay is based on the detection of differences in PCR-fragment length using capillary electrophoresis. We discuss technical problems, solutions, and the steps taken for validation. We employed the novel assay to investigate the '523' allele frequencies in different ethnicities. Whites and Hispanics have similar frequencies of S/L/VL alleles (0.45/0.11/0.44 and 0.43/0.09/0.48, respectively). In African-Americans, the frequency of the L-allele (0.10) is similar to Whites and Hispanics; however, the S-allele is more prevalent (0.65) and the VL-allele is concomitantly less frequent (0.25). The allele frequencies determined using the new methodology are compared to previous reports for Ghanaian, Japanese, Korean and Han Chinese cohorts. Finally, we studied the linkage pattern between TOMM40-'523' and APOE alleles. In Whites and Hispanics, consistent with previous reports, the L is primarily linked to ε4, while the majority of the VL and S are linked to ε3. Interestingly, in African-Americans, Ghanaians and Japanese, there is an increased frequency of the '523'S-APOEε4 haplotype. These data may be used as references for '523' allele and '523'-APOE haplotype frequencies in diverse populations for the design of research studies and clinical trials.

The Alzheimer's associated 5' region of the SORL1 gene cis regulates SORL1 transcripts expression.

SORL1 has been identified as a major contributor to late onset Alzheimer's disease (LOAD). We test whether genetic variability in the 5' of SORL1 gene modulates the risk to develop LOAD via regulation of SORL1-messenger ribonucleic acid (mRNA) expression and splicing. Two brain structures, differentially vulnerable to LOAD pathology, were examined in 144 brain samples from 92 neurologically normal individuals. The temporal cortex, which is more susceptible to Alzheimer's pathology, demonstrated ∼2-fold increase in SORL1-mRNA levels in carriers of the minor alleles at single nucleotide polymorphisms (SNPs), rs7945931 and rs2298525, compared with noncarriers. No genetic effect on total-SORL1-mRNA levels was detected in the frontal cortex. However, rs11600875 minor allele was associated with significantly increased levels of exon-2 skipping, but only in frontal cortex. No correlation of SORL1-mRNAs expression was found between frontal and temporal cortexes. Collectively, these indicate the brain region specificity of the genetic regulation of SORL1 expression. Our results suggest that genetic regulation of SORL1 expression plays a role in disease risk and may be responsible for the reported LOAD associations. Further studies to detect the actual pathogenic variant/s are necessary.

Thrombospondin-4 polymorphism (A387P) predicts cardiovascular risk in postinfarction patients with high HDL cholesterol and C-reactive protein levels.

Few studies are available in human populations investigating involvement of vascular inflammation and oxidative stress-related dysfunctional transformation of high-density lipoprotein (HDL) in establishing cardiovascular disease (CVD) risk. To this end, the current work investigated a subgroup of post-infarction patients at high-risk for recurrent events defined by high levels of HDL cholesterol (HDL-C) and concurrently high levels of C-reactive protein (CRP). Thrombospondin-4 (TSP-4), a matricellular protein of vessel walls associated with inflammation, was investigated in terms of CVD risk using multivariable modelling with a well-characterised functional genetic polymorphism of THBS4 (A387P, rs1866389) along with previously demonstrated risk-related functional genetic polymorphisms of CYBA (C242T, rs4673) and CETP (TaqIB, rs708272), and a set of blood markers. Results revealed risk-association for the gain-of-function P-allele of the THBS4 polymorphism (hazard ratio 2.00, 95% confidence interval 1.10-3.65, p=0.024). Additionally, von Willebrand factor was associated with D-dimer levels in the higher-risk P allele patients suggestive of a connection between endothelial dysfunction and thrombogenesis. In conclusion, TSP-4, a matricellular protein involved in regulating vascular inflammation, plays a role in establishing recurrent coronary risk in post-infarction patients with high levels of HDL-C and CRP. Further studies should focus on additional effects of vascular inflammatory processes on anti-atherogenic functionality of HDL particles.

Characterization of serum proteins associated with IL28B genotype among patients with chronic hepatitis C.

INTRODUCTION: Polymorphisms near the IL28B gene (e.g. rs12979860) encoding interferon λ3 have recently been associated with both spontaneous clearance and treatment response to pegIFN/RBV in chronic hepatitis C (CHC) patients. The molecular consequences of this genetic variation are unknown. To gain further insight into IL28B function we assessed the association of rs12979860 with expression of protein quantitative traits (pQTL analysis) generated using open-platform proteomics in serum from patients. METHODS: 41 patients with genotype 1 chronic hepatitis C infection from the Duke Liver Clinic were genotyped for rs12979860. Proteomic profiles were generated by LC-MS/MS analysis following immunodepletion of serum with MARS14 columns and trypsin-digestion. Next, a latent factor model was used to classify peptides into metaproteins based on co-expression and using only those peptides with protein identifications. Metaproteins were then analyzed for association with IL28B genotype using one-way analysis of variance. RESULTS: There were a total of 4,186 peptides in the data set with positive identifications. These were matched with 253 proteins of which 110 had two or more associated, identified peptides. The IL28B treatment response genotype (rs12979860_CC) was significantly associated with lower serum levels of corticosteroid binding globulin (CBG; p = 9.2×10(-6)), a major transport protein for glucocorticoids and progestins. Moreover, the CBG metaprotein was associated with treatment response (p = 0.0148), but this association was attenuated when both IL28B genotype and CBG were included in the model, suggesting that the CBG association may be independent of treatment response. CONCLUSIONS: In this cohort of chronic hepatitis C patients, IL28B polymorphism was associated with serum levels of corticosteroid binding globulin, a major transporter of cortisol, however, CBG does not appear to mediate the association of IL28B with treatment response. Further investigation of this pathway is warranted to determine if it plays a role in other comorbidities of HCV-infection.

Beneficial IL28B genotype associated with lower frequency of hepatic steatosis in patients with chronic hepatitis C.

BACKGROUND & AIMS: IL28B polymorphisms have been associated with both treatment induced and spontaneous clearance of hepatitis C virus (HCV). We previously found that LDL cholesterol levels were higher in chronic hepatitis C (CHC) patients with the CC genotype at the rs12979860 polymorphism, located proximal to the IL28 gene. Here we analyzed the association of steatosis with IL28B genotype in treatment naïve patients with CHC. METHODS: Two independent cohorts of 145 genotype 1 infected patients from an antifibrotic study and 180 genotype 1 patients from Duke were analyzed for the presence and severity of steatosis in relation to the rs12979860 polymorphism at the IL28B locus. TaqMan assay based genotyping classified three groups CC, CT, and TT. RESULTS: CC genotype was associated with a lower prevalence of steatosis. In the antifibrotic study, steatosis was found in 47.6% (50/105) of IL28B non-CC vs. 22.5% (9/40; p=0.008) in CC patients. Similarly, steatosis was found in 67.4% (89/132) of non-CC patients compared to only 39.6% (19/48; p=0.001) of CC patients in the Duke cohort. CONCLUSIONS: IL28B CC genotype is associated with less pronounced disturbances of lipid metabolism, as reflected both in serum lipoprotein levels and hepatic steatosis, in HCV infection.

High predictive accuracy of an unbiased proteomic profile for sustained virologic response in chronic hepatitis C patients.

UNLABELLED: Chronic hepatitis C (CHC) infection is a leading cause of endstage liver disease. Current standard-of-care (SOC) interferon-based therapy results in sustained virological response (SVR) in only one-half of patients, and is associated with significant side effects. Accurate host predictors of virologic response are needed to individualize treatment regimens. We applied a label-free liquid chromatography mass spectrometry (LC-MS)-based proteomics discovery platform to pretreatment sera from a well-characterized and matched training cohort of 55 CHC patients, and an independent validation set of 41 CHC genotype 1 patients with characterized IL28B genotype. Accurate mass and retention time methods aligned samples to generate quantitative peptide data, with predictive modeling using Bayesian sparse latent factor regression. We identified 105 proteins of interest with two or more peptides, and a total of 3,768 peptides. Regression modeling selected three identified metaproteins, vitamin D binding protein, alpha 2 HS glycoprotein, and Complement C5, with a high predictive area under the receiver operator characteristic curve (AUROC) of 0.90 for SVR in the training cohort. A model averaging approach for identified peptides resulted in an AUROC of 0.86 in the validation cohort, and correctly identified virologic response in 71% of patients without the favorable IL28B "responder" genotype. CONCLUSION: Our preliminary data indicate that a serum-based protein signature can accurately predict treatment response to current SOC in most CHC patients.

Gene by sex interaction for measures of obesity in the framingham heart study.

Obesity is an increasingly prevalent and severe health concern with a substantial heritable component and marked sex differences. We sought to determine if the effect of genetic variants also differed by sex by performing a genome-wide association study modeling the effect of genotype-by-sex interaction on obesity phenotypes. Genotype data from individuals in the Framingham Heart Study Offspring cohort were analyzed across five exams. Although no variants showed genome-wide significant gene-by-sex interaction in any individual exam, four polymorphisms displayed a consistent BMI association (P-values .00186 to .00010) across all five exams. These variants were clustered downstream of LYPLAL1, which encodes a lipase/esterase expressed in adipose tissue, a locus previously identified as having sex-specific effects on central obesity. Primary effects in males were in the opposite direction from females and were replicated in Framingham Generation 3. Our data support a sex-influenced association between genetic variation at the LYPLAL1 locus and obesity-related traits.

The effect of SNCA 3' region on the levels of SNCA-112 splicing variant.

Genetic variability at the 3' region of SNCA locus has been repeatedly associated with susceptibility to sporadic Parkinson's disease (PD). Accumulated evidence emphasizes the importance of SNCA dosage and expression levels in PD pathogenesis. However, the mechanism through which the 3' region of SNCA gene modulates the risk to develop sporadic PD remained elusive. We studied the effect of PD risk-associated variants at SNCA 3' regions on SNCA112-mRNA (exon 5 in-frame skipping) levels in vivo in 117 neuropathologically normal, human brain frontal cortex samples. SNPs tagging the SNCA 3' showed significant effects on the relative levels of SNCA112-mRNA from total SNCA transcripts levels. The "risk" alleles were correlated with increased expression ratio of SNCA112-mRNA from total. We provide evidence for functional consequences of PD-associated SNCA gene variants at the 3' region, suggesting that genetic regulation of SNCA splicing plays an important role in the development of the disease. Further studies to determine the definite functional variant/s within SNCA 3'and to establish their association with PD pathology are necessary.