An efficient genotyper and star-allele caller for pharmacogenomics.

High-throughput sequencing provides sufficient means for determining genotypes of clinically important pharmacogenes that can be used to tailor medical decisions to individual patients. However, pharmacogene genotyping, also known as star-allele calling, is a challenging problem that requires accurate copy number calling, structural variation identification, variant calling, and phasing within each pharmacogene copy present in the sample. Here we introduce Aldy 4, a fast and efficient tool for genotyping pharmacogenes that uses combinatorial optimization for accurate star-allele calling across different sequencing technologies. Aldy 4 adds support for long reads and uses a novel phasing model and improved copy number and variant calling models. We compare Aldy 4 against the current state-of-the-art star-allele callers on a large and diverse set of samples and genes sequenced by various sequencing technologies, such as whole-genome and targeted Illumina sequencing, barcoded 10x Genomics, and Pacific Biosciences (PacBio) HiFi. We show that Aldy 4 is the most accurate star-allele caller with near-perfect accuracy in all evaluated contexts, and hope that Aldy remains an invaluable tool in the clinical toolbox even with the advent of long-read sequencing technologies.

CYP2C8, CYP2C9, and CYP2C19 Characterization Using Next-Generation Sequencing and Haplotype Analysis: A GeT-RM Collaborative Project.

Pharmacogenetic tests typically target selected sequence variants to identify haplotypes that are often defined by star (∗) allele nomenclature. Due to their design, these targeted genotyping assays are unable to detect novel variants that may change the function of the gene product and thereby affect phenotype prediction and patient care. In the current study, 137 DNA samples that were previously characterized by the Genetic Testing Reference Material (GeT-RM) program using a variety of targeted genotyping methods were recharacterized using targeted and whole genome sequencing analysis. Sequence data were analyzed using three genotype calling tools to identify star allele diplotypes for CYP2C8, CYP2C9, and CYP2C19. The genotype calls from next-generation sequencing (NGS) correlated well to those previously reported, except when novel alleles were present in a sample. Six novel alleles and 38 novel suballeles were identified in the three genes due to identification of variants not covered by targeted genotyping assays. In addition, several ambiguous genotype calls from a previous study were resolved using the NGS and/or long-read NGS data. Diplotype calls were mostly consistent between the calling algorithms, although several discrepancies were noted. This study highlights the utility of NGS for pharmacogenetic testing and demonstrates that there are many novel alleles that are yet to be discovered, even in highly characterized genes such as CYP2C9 and CYP2C19.

Cost-effectiveness of genome sequencing for diagnosing patients with undiagnosed rare genetic diseases.

PURPOSE: To estimate the cost-effectiveness of genome sequencing (GS) for diagnosing critically ill infants and noncritically ill pediatric patients (children) with suspected rare genetic diseases from a United States health sector perspective. METHODS: A decision-analytic model was developed to simulate the diagnostic trajectory of patients. Parameter estimates were derived from a targeted literature review and meta-analysis. The model simulated clinical and economic outcomes associated with 3 diagnostic pathways: (1) standard diagnostic care, (2) GS, and (3) standard diagnostic care followed by GS. RESULTS: For children, costs of GS ($7284) were similar to that of standard care ($7355) and lower than that of standard care followed by GS pathways ($12,030). In critically ill infants, when cost estimates were based on the length of stay in the neonatal intensive care unit, the lowest cost pathway was GS ($209,472). When only diagnostic test costs were included, the cost per diagnosis was $17,940 for standard, $17,019 for GS, and $20,255 for standard care followed by GS. CONCLUSION: The results of this economic model suggest that GS may be cost neutral or possibly cost saving as a first line diagnostic tool for children and critically ill infants.

The diagnostic trajectory of infants and children with clinical features of genetic disease.

We characterized US pediatric patients with clinical indicators of genetic diseases, focusing on the burden of disease, utilization of genetic testing, and cost of care. Curated lists of diagnosis, procedure, and billing codes were used to identify patients with clinical indicators of genetic disease in healthcare claims from Optum's de-identified Clinformatics® Database (13,076,038 unique patients). Distinct cohorts were defined to represent permissive and conservative estimates of the number of patients. Clinical phenotypes suggestive of genetic diseases were observed in up to 9.4% of pediatric patients and up to 44.7% of critically-ill infants. Compared with controls, patients with indicators of genetic diseases had higher utilization of services (e.g., mean NICU length of stay of 31.6d in a cohort defined by multiple congenital anomalies or neurological presentations compared with 10.1d for patients in the control population (P < 0.001)) and higher overall costs. Very few patients received any genetic testing (4.2-8.4% depending on cohort criteria). These results highlight the substantial proportion of the population with clinical features associated with genetic disorders and underutilization of genetic testing in these populations.

Cyrius: accurate CYP2D6 genotyping using whole-genome sequencing data.

Responsible for the metabolism of ~21% of clinically used drugs, CYP2D6 is a critical component of personalized medicine initiatives. Genotyping CYP2D6 is challenging due to sequence similarity with its pseudogene paralog CYP2D7 and a high number and variety of common structural variants (SVs). Here we describe a novel bioinformatics method, Cyrius, that accurately genotypes CYP2D6 using whole-genome sequencing (WGS) data. We show that Cyrius has superior performance (96.5% concordance with truth genotypes) compared to existing methods (84-86.8%). After implementing the improvements identified from the comparison against the truth data, Cyrius's accuracy has since been improved to 99.3%. Using Cyrius, we built a haplotype frequency database from 2504 ethnically diverse samples and estimate that SV-containing star alleles are more frequent than previously reported. Cyrius will be an important tool to incorporate pharmacogenomics in WGS-based precision medicine initiatives.

PARC report: a health-systems focus on reimbursement and patient access to pharmacogenomics testing.

Pharmacogenomics test coverage and reimbursement are major obstacles to clinical uptake. Several early adopter programs have been successfully initiated through dedicated investments by federal and institutional research funding. As a result of research endeavors, evidence has grown sufficiently to support development of pharmacogenomics guidelines. However, clinical uptake is still limited. Third-party payer support plays an important role in increasing adoption, which to date has been limited to reactive single-gene testing. Access to and interest in direct-to-consumer genetic testing are driving demand for increasing healthcare providers and third-party awareness of this burgeoning field. Pharmacogenomics implementation models developed by early adopters promise to expand patient access and options, as testing continues to increase due to growing consumer interest and falling test prices.

Three patients with homozygous familial hypercholesterolemia: Genomic sequencing and kindred analysis.

BACKGROUND: Homozygous Familial Hypercholesterolemia (HoFH) is an inherited recessive condition associated with extremely high levels of low-density lipoprotein (LDL) cholesterol in affected individuals. It is usually caused by homozygous or compound heterozygous functional mutations in the LDL receptor (LDLR). A number of mutations causing FH have been reported in literature and such genetic heterogeneity presents great challenges for disease diagnosis. OBJECTIVE: We aim to determine the likely genetic defects responsible for three cases of pediatric HoFH in two kindreds. METHODS: We applied whole exome sequencing (WES) on the two probands to determine the likely functional variants among candidate FH genes. We additionally applied 10x Genomics (10xG) Linked-Reads whole genome sequencing (WGS) on one of the kindreds to identify potentially deleterious structural variants (SVs) underlying HoFH. A PCR-based screening assay was also established to detect the LDLR structural variant in a cohort of 641 patients with elevated LDL. RESULTS: In the Caucasian kindred, the FH homozygosity can be attributed to two compound heterozygous LDLR damaging variants, an exon 12 p.G592E missense mutation and a novel 3kb exon 1 deletion. By analyzing the 10xG phased data, we ascertained that this deletion allele was most likely to have originated from a Russian ancestor. In the Mexican kindred, the strikingly elevated LDL cholesterol level can be attributed to a homozygous frameshift LDLR variant p.E113fs. CONCLUSIONS: While the application of WES can provide a cost-effective way of identifying the genetic causes of FH, it often lacks sensitivity for detecting structural variants. Our finding of the LDLR exon 1 deletion highlights the broader utility of Linked-Read WGS in detecting SVs in the clinical setting, especially when HoFH patients remain undiagnosed after WES.

Correction: Estimating the burden and economic impact of pediatric genetic disease.

This Article was originally published under Nature Research's License to Publish, but has now been made available under a [CC BY-NC-ND 4.0] license. The PDF and HTML versions of the Article have been modified accordingly.

Estimating the burden and economic impact of pediatric genetic disease.

PURPOSE: To identify the economic impact of pediatric patients with clinical indications of genetic disease (GD) on the US health-care system. METHODS: Using the 2012 Kids' Inpatient Database, we identified pediatric inpatient discharges with International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes linked to genetic disease, including well-established genetic disorders, neurological diseases, birth defects, and other physiological or functional abnormalities with a genetic basis. Cohort characteristics and health-care utilization measures were analyzed. Discharges with a GD-associated primary diagnosis were used to estimate the minimum burden; discharges with GD-associated primary or secondary codes established the maximum burden. RESULTS: Of 5.85 million weighted discharges, 2.6-14% included GD-associated ICD-9-CM codes. For these discharges, mean total costs were $16,000-77,000 higher (P < 0.0001) in neonates and $12,000-17,000 higher (P < 0.0001) in pediatric patients compared with background, corresponding to significantly higher total charges and lengths of stay. Aggregate total charges for suspected GD accounted for $14 to $57 billion (11-46%) of the "national bill" for pediatric patients in 2012. CONCLUSION: Pediatric inpatients with diagnostic codes linked to genetic disease have a significant and disproportionate impact on resources and costs in the US health-care system.

HGV2012: leveraging next-generation technology and large datasets to advance disease research.

The 13th International Meeting on Human Genome Variation and Complex Genome Analysis (HGV2012: Shanghai, China, 6th-8th September 2012) was a stimulating workshop where researchers from academia and industry explored the latest progress, challenges, and opportunities in genome variation research. Key themes included advancements in next-generation sequencing (NGS) technology, investigation of common and rare diseases, employing NGS in the clinic, utilizing large datasets that leverage biobanks and population-specific cohorts, and exploration of genomic features.

Modeling bone marrow toxicity using kinase structural motifs and the inhibition profiles of small molecular kinase inhibitors.

The cellular function of kinases combined with the difficulty of designing selective small molecule kinase inhibitors (SMKIs) poses a challenge for drug development. The late-stage attrition of SMKIs could be lessened by integrating safety information of kinases into the lead optimization stage of drug development. Herein, a mathematical model to predict bone marrow toxicity (BMT) is presented which enables the rational design of SMKIs away from this safety liability. A specific example highlights how this model identifies critical structural modifications to avoid BMT. The model was built using a novel algorithm, which selects 19 representative kinases from a panel of 277 based upon their ATP-binding pocket sequences and ability to predict BMT in vivo for 48 SMKIs. A support vector machine classifier was trained on the selected kinases and accurately predicts BMT with 74% accuracy. The model provides an efficient method for understanding SMKI-induced in vivo BMT earlier in drug discovery.

The MicroArray Quality Control (MAQC)-II study of common practices for the development and validation of microarray-based predictive models.

Gene expression data from microarrays are being applied to predict preclinical and clinical endpoints, but the reliability of these predictions has not been established. In the MAQC-II project, 36 independent teams analyzed six microarray data sets to generate predictive models for classifying a sample with respect to one of 13 endpoints indicative of lung or liver toxicity in rodents, or of breast cancer, multiple myeloma or neuroblastoma in humans. In total, >30,000 models were built using many combinations of analytical methods. The teams generated predictive models without knowing the biological meaning of some of the endpoints and, to mimic clinical reality, tested the models on data that had not been used for training. We found that model performance depended largely on the endpoint and team proficiency and that different approaches generated models of similar performance. The conclusions and recommendations from MAQC-II should be useful for regulatory agencies, study committees and independent investigators that evaluate methods for global gene expression analysis.

Identification of a kinase profile that predicts chromosome damage induced by small molecule kinase inhibitors.

Kinases are heavily pursued pharmaceutical targets because of their mechanistic role in many diseases. Small molecule kinase inhibitors (SMKIs) are a compound class that includes marketed drugs and compounds in various stages of drug development. While effective, many SMKIs have been associated with toxicity including chromosomal damage. Screening for kinase-mediated toxicity as early as possible is crucial, as is a better understanding of how off-target kinase inhibition may give rise to chromosomal damage. To that end, we employed a competitive binding assay and an analytical method to predict the toxicity of SMKIs. Specifically, we developed a model based on the binding affinity of SMKIs to a panel of kinases to predict whether a compound tests positive for chromosome damage. As training data, we used the binding affinity of 113 SMKIs against a representative subset of all kinases (290 kinases), yielding a 113x290 data matrix. Additionally, these 113 SMKIs were tested for genotoxicity in an in vitro micronucleus test (MNT). Among a variety of models from our analytical toolbox, we selected using cross-validation a combination of feature selection and pattern recognition techniques: Kolmogorov-Smirnov/T-test hybrid as a univariate filter, followed by Random Forests for feature selection and Support Vector Machines (SVM) for pattern recognition. Feature selection identified 21 kinases predictive of MNT. Using the corresponding binding affinities, the SVM could accurately predict MNT results with 85% accuracy (68% sensitivity, 91% specificity). This indicates that kinase inhibition profiles are predictive of SMKI genotoxicity. While in vitro testing is required for regulatory review, our analysis identified a fast and cost-efficient method for screening out compounds earlier in drug development. Equally important, by identifying a panel of kinases predictive of genotoxicity, we provide medicinal chemists a set of kinases to avoid when designing compounds, thereby providing a basis for rational drug design away from genotoxicity.

A meta-analysis of QTL for diabetes-related traits in rodents.

Crossbreeding studies in rodents have identified numerous quantitative trait loci (QTL) that are linked to diabetes-related component traits. To identify genetic consensus regions implicated in insulin action and glucose homeostasis, we have performed a meta-analysis of genomewide linkage scans for diabetes-related traits. From a total of 43 published genomewide scans we assembled a nonredundant collection of 153 QTL for glucose levels, insulin levels, and glucose tolerance. Collectively, these studies include data from 48 different parental strains and >11,000 individual animals. The results of the studies were analyzed by the truncated product method (TPM). The analysis revealed significant evidence for linkage of glucose levels, insulin levels, and glucose tolerance to 27 different segments of the mouse genome. The most prominent consensus regions [localized to chromosomes 2, 4, 7, 9, 11, 13, and 19; logarithm of odds (LOD) scores 10.5-17.4] cover approximately 11% of the mouse genome and collectively contain the peak markers for 47 QTL. Approximately half of these genomic segments also show significant linkage to body weight and adiposity, indicating the presence of multiple obesity-dependent and -independent consensus regions for diabetes-related traits. At least 84 human genetic markers from genomewide scans and >80 candidate genes from human and rodent studies map into the mouse consensus regions for diabetes-related traits, indicating a substantial overlap between the species. Our results provide guidance for the identification of novel candidate genes and demonstrate the presence of numerous distinct consensus QTL regions with highly significant LOD scores that control glucose homeostasis. An interactive physical map of the QTL is available online at http://www.diabesitygenes.org.