Variability in GWAS analysis: the impact of genotype calling algorithm inconsistencies.

The Genome-Wide Association Working Group (GWAWG) is part of a large-scale effort by the MicroArray Quality Consortium (MAQC) to assess the quality of genomic experiments, technologies and analyses for genome-wide association studies (GWASs). One of the aims of the working group is to assess the variability of genotype calls within and between different genotype calling algorithms using data for coronary artery disease from the Wellcome Trust Case Control Consortium (WTCCC) and the University of Ottawa Heart Institute. Our results show that the choice of genotyping algorithm (for example, Bayesian robust linear model with Mahalanobis distance classifier (BRLMM), the corrected robust linear model with maximum-likelihood-based distances (CRLMM) and CHIAMO (developed and implemented by the WTCCC)) can introduce marked variability in the results of downstream case-control association analysis for the Affymetrix 500K array. The amount of discordance between results is influenced by how samples are combined and processed through the respective genotype calling algorithm, indicating that systematic genotype errors due to computational batch effects are propagated to the list of single-nucleotide polymorphisms found to be significantly associated with the trait of interest. Further work using HapMap samples shows that inconsistencies between Affymetrix arrays and calling algorithms can lead to genotyping errors that influence downstream analysis.

Batch effects in the BRLMM genotype calling algorithm influence GWAS results for the Affymetrix 500K array.

The Affymetrix GeneChip Human Mapping 500K array is common for genome-wide association studies (GWASs). Recent findings highlight the importance of accurate genotype calling algorithms to reduce the inflation in Type I and Type II error rates. Differential results due to genotype calling errors can introduce severe bias in case-control association study results. Using data from the Wellcome Trust Case Control Consortium, 1991 individuals with coronary artery disease (CAD) and 1500 controls from the UK Blood Services (NBS) were genotyped on the Affymetrix 500K array. Different batch sizes and compositions were used in the Bayesian Robust Linear Model with Mahalanobis distance classifier (BRLMM) genotype calling algorithm to assess the batch effect on downstream association analysis. Results show that composition (cases and controls genotyped simultaneously or separate) and size (number of individuals processed by BRLMM at a time) can create 2-3% discordance in the results for quality control and statistical analysis and may contribute to the lack of reproducibility between GWASs. The changes in batch size are largely responsible for differential single-nucleotide polymorphism results, yet we observe evidence of an interactive effect of batch size and composition that contributes to discordant results in the list of significantly associated loci.

Assessment of variability in GWAS with CRLMM genotyping algorithm on WTCCC coronary artery disease.

The robustness of genome-wide association study (GWAS) results depends on the genotyping algorithms used to establish the association. This paper initiated the assessment of the impact of the Corrected Robust Linear Model with Maximum Likelihood Classification (CRLMM) genotyping quality on identifying real significant genes in a GWAS with large sample sizes. With microarray image data from the Wellcome Trust Case-Control Consortium (WTCCC), 1991 individuals with coronary artery disease (CAD) and 1500 controls, genetic associations were evaluated under various batch sizes and compositions. Experimental designs included different batch sizes of 250, 350, 500, 2000 samples with different distributions of cases and controls in each batch with either randomized or simply combined (4:3 case-control ratios) or separate case-control samples as well as whole 3491 samples. The separate composition could create 2-3% discordance in the single nucleotide polymorphism (SNP) results for quality control/statistical analysis and might contribute to the lack of reproducibility between GWAS. CRLMM shows high genotyping accuracy and stability to batch effects. According to the genotypic and allelic tests (P<5.0 x 10(-7)), nine significant signals on chromosome 9 were found consistently in all batch sizes with combined design. Our findings are critical to optimize the reproducibility of GWAS and confirm the genetic role in the pathophysiology of CAD.

The Critical Path Institute's approach to precompetitive sharing and advancing regulatory science.

Many successful large industries, such as computer-chip manufacturers, the cable television industry, and high-definition television developers,(1) have established successful precompetitive collaborations focusing on standards, applied science, and technology that advance the field for all stakeholders and benefit the public.(2) The pharmaceutical industry, however, has a well-earned reputation for fierce competition and did not demonstrate willingness to share data or knowledge until the US Food and Drug Administration (FDA) launched the Critical Path Initiative in 2004 (ref. 3).

Assessing sources of inconsistencies in genotypes and their effects on genome-wide association studies with HapMap samples.

The discordance in results of independent genome-wide association studies (GWAS) indicates the potential for Type I and Type II errors. We assessed the repeatibility of current Affymetrix technologies that support GWAS. Reasonable reproducibility was observed for both raw intensity and the genotypes/copy number variants. We also assessed consistencies between different SNP arrays and between genotype calling algorithms. We observed that the inconsistency in genotypes was generally small at the specimen level. To further examine whether the differences from genotyping and genotype calling are possible sources of variation in GWAS results, an association analysis was applied to compare the associated SNPs. We observed that the inconsistency in genotypes not only propagated to the association analysis, but was amplified in the associated SNPs. Our studies show that inconsistencies between SNP arrays and between genotype calling algorithms are potential sources for the lack of reproducibility in GWAS results.

Novel biomarkers of acute kidney toxicity.

Novel biomarkers of kidney toxicity are powerful tools not only with respect to their clinical applications but also because of their impact on drug development. These biomarkers can influence the assessment of efficacy of new drugs for kidney diseases as well as the risk management for new drugs. The science behind these novel biomarkers reflects the evolution over the past decade of genomic and proteomic platforms that have transformed the discovery and development of new biomarkers for preclinical and clinical applications in drug development. Several of these biomarkers are in use as transcriptomic biomarkers in animal models as well as translational proteomic biomarkers in animal models and in humans. Their ability to detect kidney damage earlier than is possible with currently accessible biomarkers is being given qualification through regulatory biomarker-qualification programs, which will help establish consensus for their widespread use.

The experience with voluntary genomic data submissions at the FDA and a vision for the future of the voluntary data submission program.

Drug developers have been using genomic information in drug development strategies for a number of years, but it was unclear how this information would be reviewed by the Food and Drug Administration (FDA). In order to evaluate the regulatory impact of genomic data in current drug development, a workshop was held in May 2002 to discuss aspects surrounding genomic data submission to the FDA (Figure 1).

A novel method to compensate for different amplification efficiencies between patient DNA samples in quantitative real-time PCR.

Quantification of residual disease by real-time polymerase chain reaction (PCR) will become a pivotal tool in the development of patient-directed therapy. In recent years, various protocols to quantify minimal residual disease in leukemia or lymphoma patients have been developed. These assays assume that PCR efficiencies are equal for all samples. Determining t(14;18) and albumin reaction efficiencies for sixteen follicular lymphoma patient samples revealed higher efficiencies for blood samples than for lymph node samples in general. However, within one sample both reactions had equivalent efficiencies. Differences in amplification efficiencies between patient samples (low efficiencies) and the calibrator in quantitative analyses result in the underestimation of residual disease in patient samples whereby the weakest positive patient samples are at highest error. Based on these findings for patient samples, the efficiency compensation control was developed. This control includes two reference reactions in a multiplex setting, specific for the beta-actin and albumin housekeeping genes that are present in a constant ratio within DNA templates. The difference in threshold cycle values for both reference reactions, ie, the Ct(2) value, is dependent on the amplification efficiency, and is used to compensate for efficiency differences between patient samples and the calibrator. The beta-actin reference reaction is also used to normalize for DNA input. Furthermore, the efficiency compensation control facilitates identification of patient samples that are so contaminated with PCR inhibitory compounds that different amplification reactions are affected to a different extent. Accurate quantitation of residual disease in these samples is therefore impossible with the current quantitative real-time PCR protocols. Identification and exclusion of these inadequate samples will be of utmost importance in quantitative retrospective studies, but even more so, in future molecular diagnostic analyses.

5' Nuclease assays for the loci CCR5-+/Delta32, CCR2-V64I, and SDF1-G801A related to pathogenesis of AIDS.

BACKGROUND: Variations within the human genome play important roles in human disease. To study variations related to susceptibility to AIDS, we have developed 5' nuclease assays that eliminate post-PCR molecular biology steps. METHODS: TaqMan assays based on the 5' nuclease activity of Taq polymerase and fluorescent resonance energy transfer were developed to score alleles at the biallelic loci CCR5-+/Delta32, CCR2-V64I and SDF1-G801A. For each assay, 72 samples were analyzed. Data collection and analysis were performed on the Prism 7700 Sequence Detection System. For comparison with gel electrophoresis methods, each locus was also scored on a subset of 24 samples, using restriction enzymes or single-strand conformational polymorphism (SSCP). RESULTS: Clear allelic discrimination was obtained on each of the 72 samples for all three TaqMan assays. The TaqMan scores for the subset of 24 samples were concordant with the restriction enzyme and SSCP scores. CONCLUSIONS: Because of its simplicity, speed, and potential for automation and miniaturization, TaqMan is an excellent candidate for investigation of genetic variation in clinical, research, and forensic settings.

Differential cytokine and chemokine gene expression by human NK cells following activation with IL-18 or IL-15 in combination with IL-12: implications for the innate immune response.

NK cells constitutively express monocyte-derived cytokine (monokine) receptors and secrete cytokines and chemokines following monokine stimulation, and are therefore a critical component of the innate immune response to infection. Here we compared the effects of three monokines (IL-18, IL-15, and IL-12) on human NK cell cytokine and chemokine production. IL-18, IL-15, or IL-12 alone did not stimulate significant cytokine or chemokine production in resting NK cells. The combination of IL-18 and IL-12 induced extremely high amounts of IFN-gamma protein (225 +/- 52 ng/ml) and a 1393 +/- 643-fold increase in IFN-gamma gene expression over those in resting NK cells. IL-15 and IL-12 induced less IFN-gamma protein (24 +/- 10 ng/ml; p < 0.007) and only a 45 +/- 19-fold increase in IFN-gamma gene expression over those in resting NK cells. The CD56bright NK cell subset produced significantly more IFN-gamma following IL-18 and IL-12 compared with CD56dim NK cells (p < 0.008). However, the combination of IL-15 and IL-12 was significantly more potent than that of IL-18 and IL-12 for NK cell production of IL-10, macrophage inflammatory protein-1alpha, macrophage inflammatory protein-1beta, and TNF-alpha at the protein and transcript levels. Granulocyte-macrophage CSF was optimally induced by IL-15 and IL-18. Resting CD56+ NK cells expressed IL-18R transcript that was up-regulated by IL-12 or IL-15. Our results show that distinct cytokine and chemokine patterns are induced in NK cells in response to different costimulatory signals from these three monokines. This suggests that NK cell cytokine production may be governed in part by the monokine milieu induced during the early proinflammatory response to infection and by the subset of NK cells present at the site of inflammation.

Automated detection of prevalent mutations in BRCA1 and BRCA2 genes, using a fluorogenic PCR allelic discrimination assay.

Mutations in the genes BRCA1 and BRCA2 account for 5%-10% of familial early onset breast cancer. Identification of these mutations allows molecular diagnosis for breast cancer susceptibility. A high through-put automated PCR allelic discrimination assay (ADA) was developed to detect the prevalent mutations in these genes. Two allele specific oligonucleotides (ASO) are directly used in the PCR reaction, in both of which the fluorescent reporter and quencher dyes are attached to the 5' and 3' ends, respectively. During PCR, fluorescence is generated after cleavage of the annealed ASO by the 5' nuclease activity of Taq polymerase. The wild-type BRCA sequence is distinguished from the mutant sequence by the differential fluorescence emission of two different reporter dyes. The sensitivity of ADA is at the level of a single cell following a nested PCR. Eighty-six patient samples can be analyzed for each mutation in 15-min post-PCR without the need for radioactivity, gel electrophoresis, or membrane blotting/hybridization.

Study of Fc gamma murine macrophage receptors (Fc gamma R) in the internalization of Histoplasma capsulatum yeasts.

Reports about Fc gamma R involvement in the phagocytosis of Histoplasma capsulatum yeast are contradictory. Fungal cells were incubated with anti-Histoplasma murine IgG or normal IgG as control. The opsonized yeasts were adjusted and spread on a J774.2 macrophage cell line monolayer. Yeast cells were allowed to attach to the monolayers at 4 degrees C for 1 h prior to initiating the internalization step of the phagocytic process, by shifting the cells to 37 degrees C. The participation of Fc gamma R in the internalization of yeast was monitored by radioiodination labelling of the macrophage surface proteins, and rosette-forming cells for erythrocytes coated with antibodies (RFC-EA). Results showed decreased [125I]-incorporation on the macrophage surface proteins at 15 min post-anti-Histoplasma IgG opsonized yeast infection. The same was observed when high and low virulent H. capsulatum strains were used. A decrease of RFC percentage was also observed at 15 min after IgG opsonized yeast phagocytosis. Results support the disappearance of Fc gamma R at 15 min from the surface of the macrophage that had ingested anti-Histoplasma IgG opsonized yeasts. The recovery of [125I]-counts and the RFC response at a longer time of phagocytosis suggest a possible Fc gamma R recycling mechanism.

Evaluation of phagolysosome fusion in acridine orange stained macrophages infected with Histoplasma capsulatum.

A phagosome-lysosome (PL) fusion was performed in vitro using peritoneal cells from normal BALB/c mice and the J774.2 macrophage cell line infected with the yeast phase of the fungus Histoplasma capsulatum at ratios of 5 x 10(5), 5 x 10(6) or 1 x 10(7) yeasts per 1 x 10(6) macrophages, and phagocytosis was allowed to proceed for 5, 30 and 60 min. Macrophage lysosomes were pre-labelled with acridine orange and the cells were challenged with the parasite. Fusion was evaluated by fluorescence microscopy and the number of macrophages with stained yeast cells was scored. The phagolysosome fusion frequency (PLFF) was calculated by subtracting the specific fusions of infected macrophages from the non-specific fusions of uninfected macrophages and normalizing the total number of bound yeasts. The PLFF was determined using different doses and strains of H. capsulatum. Results showed that PLFF in infected macrophages depends on the infection dose. Inhibition of PL fusion was detected mainly at a high infection ratio (1 x 10(7) yeasts/1 x 10(6) macrophages), and PL fusion varied with phagocytosis time. No significant differences were observed in the fusions when different Histoplasma strains were used. Results with J774.2 cells were similar to peritoneal cells, indicating that both methodology and fusion calculations employed were useful, in spite of the heterogeneity of macrophage populations used.

Modulation of macrophage lysosomal pH by Mycobacterium tuberculosis-derived proteins.

Macrophage lysosomal pH was significantly (greater than 1 pH unit) increased in a reversible, concentration-dependent manner characterized by a saturable and cyclic kinetics after exposure to culture filtrate protein extract derived from Mycobacterium tuberculosis. Lysosomal alkalinization peaked 30 min after administration of culture filtrate protein extract to cells of the macrophagelike cell line J774A.1. The alkalinization was reversible, and a second peak was observed approximately 60 min after incubation. Maximum lysosomal alkalinization increased as a function of culture filtrate protein extract concentration, reaching an apparent saturation level around 700 to 1,000 micrograms/ml, although the time course for this process was not significantly dependent on antigen concentration. The alkalinizing agent(s) was heat labile and produced a similar effect in cells which had a different lysosomal enzyme composition. Our observations are consistent with the presence of one or more mycobacterial antigens which have a pH-dependent affinity for lysosomal structures essential for lysosomal acidification and which are able to inhibit this lysosomal acidification.

Analysis of porphyrins and enzymes in porphyrin synthesis in Taenia solium cysticercus from man and pig.

Porphyrins and activities of heme biosynthetic enzymes in Taenia solium cysticerci from porcine and human hosts, were examined in order to clarify the possible step where heme synthesis is interrupted. Porphyrins in the vesicular fluid of the parasite were predominantly coproporphyrin, followed by penta-carboxylated porphyrin, which together accounted for 90% of the accumulated porphyrins. Coproporphyrin and penta-carboxylated porphyrin were both type I and III isomers. Small amounts of protoporphyrin and uroporphyrin, and trace amounts of tri-, hexa- and hepta-carboxylated porphyrins were also detected. Fluorescence and phosphorescence spectra and lifetime studies revealed that at least 75% of the porphyrins were bound to metal, probably Zn, while the rest was free. Reverse phase high performance liquid chromatography monitored at an excitation wavelength of 417 nm and at an emission wavelength of 585 nm demonstrated that approximately 90% of these porphyrins were Zn-coproporphyrin. A fluorescence excitation peak at 283 nm with an emission peak at 585 nm and 625 nm indicated that some of the porphyrins were associated with proteins in the vesicular fluid of the parasite. Low levels of delta-aminolevulinic acid dehydratase, porphobilinogen deaminase and uroporphyrinogen decarboxylase activities, and heme concentrations were found in the extract of the parasite walls and scolex, but not in the vesicular fluid. The porphyrin accumulation pattern in this parasite can best be explained by postulating a deficiency of coproporphyrinogen oxidase activity, similar to that in human patients with hereditary coproporphyria. A parasite dissected from a human host was considerably less porphyric than those from pigs, but the pattern of accumulated porphyrins was quite similar in both. In view of their porphyrin contents, T. solium cysticerci could be light sensitive.

Porphyrin content of the cysticercus of Taenia solium.

The strong red fluorescence of the cysticercus of Taenia solium depends on the presence of several porphyrins in the vesicular fluid of the parasite: probably protoporphyrin IX, coproporphyin I or III, and 2 decarboxylated porphyrins intermediate between uroporphyrin and coproporphyrin. Cyst porphyrins associated to form conglomerates of high molecular weight that dissociated in acid solutions and were not antigenic themselves nor associated with antigenic molecules. An appreciable fraction of the porphyrins was capable of undergoing oxidation and reduction, indicating that some of the porphyrins were complexed with metal ions. The metabolic basis for the accumulation of porphyrins is unknown. Preliminary results suggest that conditions deleterious to the cysticercus cause release of porphyrins so that the appearance of porphyrins in the cerebrospinal fluid of neurocysticercotic patients may prove useful in monitoring therapeutic attacks on the parasite.

Inhibition of mitogenesis induced by phytohemagglutinin and Lens culinaris lectin in adherent-cell supernatants treated with protein extract of Mycobacterium tuberculosis.

Specific stimulation of T cells by phytohemagglutinin and Lens culinaris lectin was inhibited by a soluble factor(s) secreted by normal adherent cells stimulated with culture filtrate protein extract (CFPE) derived from bacterial cultures of Mycobacterium tuberculosis H37Ra (avirulent) and H37Rv (virulent). The induction of the inhibitory factor was blocked by the presence of hyperimmune antisera to H37Rv or H37Ra CFPE. The inhibitory factor did not seem to be a CFPE reprocessed by the adherent cells. Inhibitory activity was maximal in supernatants of adherent-cell cultures incubated for 48 h; the inhibitory factor was heat labile, and its production was dependent on the concentration of M. tuberculosis CFPE. A mouse monocyte-macrophage cell line, ATCC J774A.1, produced an identical inhibitory factor, thus excluding a non-macrophage-contaminating adherent cell as the source of the factor. This inhibitory factor also interfered with the recognition of phytohemagglutinin and Lens culinaris lectin by T cells.