Adult diffuse glioma GWAS by molecular subtype identifies variants in D2HGDH and FAM20C

Abstract

Background

Twenty-five germline variants are associated with adult diffuse glioma, and some of these variants have been shown to be associated with particular subtypes of glioma. We hypothesized that additional germline variants could be identified if a genome-wide association study (GWAS) were performed by molecular subtype.

Methods

A total of 1320 glioma cases and 1889 controls were used in the discovery set and 799 glioma cases and 808 controls in the validation set. Glioma cases were classified into molecular subtypes based on combinations of isocitrate dehydrogenase (IDH) mutation, telomerase reverse transcriptase (TERT) promoter mutation, and 1p/19q codeletion. Logistic regression was applied to the discovery and validation sets to test for associations of variants with each of the subtypes. A meta-analysis was subsequently performed using a genome-wide P-value threshold of 5 × 10⁻⁸.

Results

Nine variants in or near D-2-hydroxyglutarate dehydrogenase (D2HGDH) on chromosome 2 were genome-wide significant in IDH-mutated glioma (most significant was rs5839764, meta P = 2.82 × 10⁻¹⁰). Further stratifying by 1p/19q codeletion status, one variant in D2HGDH was genome-wide significant in IDH-mutated non-codeleted glioma (rs1106639, meta P = 4.96 × 10⁻⁸). Further stratifying by TERT mutation, one variant near FAM20C (family with sequence similarity 20, member C) on chromosome 7 was genome-wide significant in gliomas that have IDH mutation, TERT mutation, and 1p/19q codeletion (rs111976262, meta P = 9.56 × 10⁻⁹). Thirty-six variants in or near GMEB2 on chromosome 20 near regulator of telomere elongation helicase 1 (RTEL1) were genome-wide significant in IDH wild-type glioma (most significant was rs4809313, meta P = 2.60 × 10⁻¹⁰).

Conclusions

Performing a GWAS by molecular subtype identified 2 new regions and a candidate independent region near RTEL1, which were associated with specific glioma molecular subtypes.

Genome-wide association studies (GWAS) identified variants in 25 regions that are associated with development of adult diffuse glioma.^1–9 Recently, GWAS by histological subtype identified novel germline variants that were associated specifically with glioblastoma (GBM; World Health Organization [WHO] grade IV) and non-GBM (grades II–III).^2,3 Importantly, these newly identified GBM and non-GBM germline variants did not reach genome-wide significance when a GWAS was performed on overall glioma; they only reached genome-wide significance when a GWAS was performed within a more homogeneous subgroup of glioma. As highlighted by the 2016 WHO classification criteria,¹⁰ which include genetic tumor testing of isocitrate dehydrogenase (IDH) mutation and 1p/19q codeletion, glioma can more accurately be subtyped by somatic alterations. Telomerase reverse transcriptase (TERT) promoter mutation has also been reproducibly shown to be associated with age at diagnosis, patient outcome, and specific germline associations.^11–14 However, a GWAS by these molecular subtypes has not been reported. Thus, we hypothesized that novel germline variants might be identified by GWAS performed by clinically relevant molecular subtypes. Here, we describe the results of performing a GWAS within molecular subtypes defined by combinations of IDH mutation, TERT promoter mutation, and 1p/19q codeletion.

Methods

Subjects

Mayo Clinic glioma cases.

Mayo Clinic glioma cases have been previously described.^3,8,11,15 The study was approved by the Mayo Clinic Office for Human Research Protection. Histologically confirmed grades II, III, and IV glioma cases were identified at diagnosis (at Mayo Clinic) or at the time of pathologic confirmation (diagnosed elsewhere and treated at Mayo Clinic). Included subjects were at least 18 years of age and had a surgical resection or biopsy between 1973 and 2014. A total of 653 cases that were run on the OncoArray genotyping assay had necessary molecular data.³

UCSF glioma cases.

UCSF glioma cases include participants of the San Francisco Bay Area Adult Glioma Study (AGS). This study was approved by the UCSF Committee on Human Research. Details of subject recruitment for AGS have been reported previously.^{1,3,8,11,16–18} Cases were adults (>18 y of age) with newly diagnosed, histologically confirmed grade II, III, or IV glioma. Population-based cases diagnosed between 1991 and 2009 and residing in the 6 San Francisco Bay Area counties were ascertained using the Cancer Prevention Institute of California’s early case ascertainment system. Clinic-based cases diagnosed between 2002 and 2012 were recruited from the UCSF Neuro-oncology Clinic, regardless of place of residence. A total of 667 cases that were run on the OncoArray genotyping assay had necessary molecular data.³

TCGA glioma cases.

A total of 799 glioma cases from The Cancer Genome Atlas (TCGA) obtained from Database of Genotypes and Phenotypes (dbGaP) (phs000178) had necessary molecular data. All samples were run on the Affymetrix 6.0 genotyping array. IDH mutation, TERT promoter mutation, and 1p/19q codeletion status were obtained from supplementary table 1 in Ceccarelli et al,¹⁹ and D-2-hydroxyglutarate dehydrogenase (D2HGDH) copy number calls were downloaded from cBioPortal on February 19, 2019.

International Glioma Case-Control (GICC) Study controls.

A total of 1907 GICC controls that were approved for “General Research Use” and “Brain Tumors” were obtained from dbGaP (phs001319); 1889 passed quality control metrics and were included in the analyses. All samples were run on the OncoArray genotyping assay.³

Mayo Clinic Biobank controls.

A total of 808 Mayo Clinic Biobank controls were utilized.²⁰ All samples were run on the Illumina Omni Express genotyping array.

Statistical Methods

GWAS analyses.

Prior to imputation, the following quality control procedures were applied to the discovery and validation sets: tests of Hardy–Weinberg equilibrium (P < 10⁻⁶), duplicate and relatedness checks, sex checks, variant call rates (>95%), and subject call rates (>95%). For the validation set, because cases and controls were run on different genotyping platforms, quality control was performed on cases and controls separately. Imputation was performed using the Michigan Imputation Server, utilizing the 1000 Genome V3 data and the Haplotype Reference Consortium. Structure²¹ was applied to determine racial groups, using 1000 Genome samples to anchor the racial groups. Population stratification was assessed using Eigenstrat and principal components (Supplementary Figure 1).²² Principal components that were significantly associated with overall glioma cases versus control (P < 0.05) were included as covariates in each of the subtype-specific GWAS analyses. Subtype-specific GWAS analyses were performed with cases defined by somatic alterations and corresponding cases being compared with controls. GWAS was first performed stratified by IDH mutation: IDH-mutated glioma and IDH wild-type glioma were each compared with controls. IDH-mutated cases were subsequently stratified by 1p/19q codeletion into IDH-mutated 1p/19q codeleted and IDH-mutated 1p/19q non-codeleted. GWAS were then conducted for subgroups further stratified by TERT promoter mutation into triple-positive (IDH-mutated, TERT mutated, and 1p/19q codeleted), IDH and TERT mutations, IDH mutation only, TERT mutation only, and triple-negative (IDH wild-type, TERT wild-type, 1p/19q non-codeleted). Logistic regression was utilized comparing subtype-specific cases with controls, with genotype coded as 0, 1, or 2 copies (or dosage if imputed) of the alternate allele, and age, sex, and principal components included as covariates. Q-Q plots and lambda values were used to evaluate the excess false-positive rate. Imputation R² value for the discovery set was required to be larger than 0.80 and the P-value threshold was 5 × 10⁻⁶. Variants that passed the imputation and P-value thresholds in the discovery set were meta-analyzed across the discovery and validation sets. Variants that had the same direction of effect in the discovery and validation sets, genome-wide significant meta P-value (P < 5 × 10⁻⁸), and P-value less than 0.05 in the validation set were further evaluated. For a subset of the imputed variants that passed genome-wide significance in the meta-analysis, 93 glioma patient samples were custom genotyped using TaqMan and compared with imputed genotypes (Supplementary Table 1).

Expression quantitative trait loci, Hi-C and ChromHMM analyses.

Expression quantitative trait loci (eQTL) analyses were performed using data obtained from the GTEx Portal on 10/22/2019. Analyses were performed for the most significant variant at each locus. All genes within 1 Mb of the variant were evaluated in normal brain tissues available in GTEx. Hi-C analyses were performed using the 3D-genome Interaction Viewer and database.²³ Interactions were examined for a 200 kb window surrounding the variant of interest in dorsolateral prefrontal cortex, hippocampus and H1-derived neural progenitor cells. ChromHMM was also evaluated for the dorsolateral prefrontal cortex and hippocampus using the UC Santa Cruz genome browser and the Roadmap Epigenomics project.²⁴

Results

GWAS by Glioma Subtypes: Subtypes Defined by IDH Mutation and 1p/19q Codeletion

Of the 2119 glioma cases analyzed in the meta-analysis, 1012 were IDH-mutated and 1107 IDH wild-type (Table 1). Further stratifying IDH-mutated glioma according to 1p/19q codeletion, in the meta-analysis 390 patients had IDH-mutated 1p/19q-codeleted tumors and 561 patients had IDH-mutated non-codeleted tumors (Table 1).

IDH-mutated GWAS.

Ninety-three variants passed genome-wide significance in IDH-mutated glioma versus controls (Fig. 1A; Supplementary Figure 2A; Supplementary Table 2). Most of the variants were in regions that have previously been reported: CCDC26, PHLDB1, AKT3, and IDH1.^1,3,5 Nine variants in or near D2HGDH on chromosome region 2q27 were genome-wide significant (Supplementary Table 2; Fig. 2A). The most significant variant in the D2HGDH region was rs5839764 (discovery odds ratio [OR] = 1.51, meta P-value = 2.82 × 10⁻¹⁰; Table 2), and this variant remained significant after adjustment for the known IDH1 variant rs7572263 on chromosome 2 (P = 5.46 × 10⁻⁷; Supplementary Table 3).³ TCGA reported that the 2q37 region was commonly deleted in IDH-mutated gliomas that do not have 1p/19q codeletion.¹⁹ In TCGA data for IDH-mutated non-codeleted glioma, we observed that rs5839764 was inversely associated with tumor deletions of D2HGDH (OR = 0.57, 95% CI: 0.36–0.90, P = 0.015), with deletions more likely to occur in patients who carry the reference allele C versus the alternative allele G. Hi-C DNA interactions were observed between rs5839764 and nearby regions, including the 5′ end of D2HGDH, in the hippocampus and H1-derived neural progenitor cells (Fig. 2A). GTEx did not have data on rs5839764, and thus the second most significant variant was evaluated as a surrogate (rs71430382). The rs71430382 alternate allele T was associated with decreased expression of D2HGDH in normal brain tissues (P = 4.9 × 10⁻¹⁴, 1.7 × 10⁻¹⁸, and 2.2 × 10⁻¹¹ for frontal cortex, cortex, and hippocampus, respectively; Fig. 2B).

IDH-mutated non-codeleted GWAS.

Twelve variants passed genome-wide significance for patients with IDH-mutated non-codeleted glioma versus controls (Fig. 1B; Supplementary Figure 2B; Supplementary Table 4). Most of the variants were in known genes, including CCDC26 and PHLDB1.^1,5 One variant within D2HGDH, rs1106639, reached genome-wide significance (discovery OR = 1.7, meta P-value = 4.96 × 10⁻⁸) (Table 2). The variant remained significantly associated with risk after adjustment for the IDH1 variant rs7572263 on chromosome 2 (P = 5.0 × 10⁻⁶; Supplementary Table 5).³ The Hi-C analyses for rs1106639 showed similar results as rs5839764 (Supplementary Figure 3).

IDH-mutated 1p/19q-codeleted GWAS.

Twenty-eight variants passed genome-wide significance in IDH-mutated codeleted glioma versus controls; all variants were in CCDC26¹ (Supplementary Figure 2C; Supplementary Figure 4; Supplementary Table 6).

IDH wild-type GWAS.

Eighty-five variants were associated with IDH wild-type glioma versus controls at genome-wide significance (Fig. 1D; Supplementary Figure 2D; Supplementary Table 7). Most of the variants were in known genes, including TERT and regulator of telomere elongation helicase 1 (RTEL1).^5,8 Thirty-six variants in or near glucocorticoid modulatory element binding protein 2 (GMEB2) on chromosome region 20q13 were genome-wide significant (Table 2; Fig. 3). The most significant variant in the GMEB2 region was rs4809313 (discovery OR = 0.66, meta P-value = 2.60 × 10⁻¹⁰), and this variant remained significant after adjustment for the known RTEL1 glioma risk variant (rs2297440) that is nearby on chromosome 20 (P = 0.029; Supplementary Table 8).^3,5 There were no significant eQTL in cerebellar hemisphere tissue. Hi-C interactions were observed between rs4809313 and nearby genes, including RTEL1 (Fig. 3). Since we observed Hi-C interactions between the GMEB2 variant and the RTEL1 region, we also evaluated Hi-C interactions with the RTEL1 glioma risk variant rs2297440. Hi-C interactions were observed between the RTEL1 variant rs2297440 and the GMEB2 region (Supplementary Figure 5).

GWAS by Glioma Subtypes: Subtypes Stratified by TERT Promoter Mutation

Further stratifying IDH-mutated tumors by TERT promoter mutation resulted in the following numbers of patients in each of the meta-analyses: 455 IDH-mutated only (IDH-mutated, TERT wild-type, and 1p/19q non-codeleted), 52 IDH- and TERT-mutated (IDH-mutated, TERT-mutated, and 1p/19q non-codeleted), and 327 triple-positive (IDH-mutated, TERT-mutated, and 1p/19q codeleted) glioma (Table 1). The meta-analyses for IDH wild-type tumors were further stratified by TERT promoter mutation into 578 TERT-mutated only (IDH wild-type, TERT-mutated, and 1p/19q non-codeleted) and 120 triple-negative (IDH wild-type, TERT wild-type, and 1p/19q non-codeleted) glioma (Table 1).

Triple-positive GWAS.

Twenty-nine variants were associated with triple-positive glioma versus controls at genome-wide significance (Fig. 1C; Supplementary Figure 6a; Supplementary Table 9). Most variants were in CCDC26¹; however, rs111976262 was in a novel region on chromosome 7 near family with sequence similarity 20, member C (FAM20C) (discovery OR = 3.52, meta P-value = 9.56 × 10⁻⁹) (Table 2). There were no significant eQTL in normal brain tissue. Hi-C interactions were observed between rs111976262 and nearby regions (Fig. 4).

IDH-mutated only GWAS.

Seven variants were associated with IDH-mutated only glioma versus controls at genome-wide significance; all variants were in CCDC26¹ and previously known (Supplementary Figure 6b; Supplementary Figure 7a; Supplementary Table 10).

IDH- and TERT-mutated GWAS.

No variants were observed to be associated at genome-wide significance with IDH- and TERT-mutated glioma versus controls (Supplementary Figure 6c; Supplementary Figure 7b).

TERT-mutated only GWAS.

Thirty-eight variants were associated with TERT-mutated only glioma at genome-wide significance (Supplementary Figure 8a; Supplementary Figure 9a; Supplementary Table 11). All variants were in or near previously established glioma risk regions including TERT and RTEL1.⁵

Triple-negative GWAS.

No variants were genome-wide significant in triple-negative glioma versus controls (Supplementary Figure 8b; Supplementary Figure 9b).

Association of D2HGDH, Allergy, and Glioma

Alternate allele A of germline variant rs34290285 within D2HGDH has been shown to be protective of allergy and asthma,^25–27 and allergy has been shown to be protective of adult diffuse glioma.^28,29 We evaluated the association of the D2HGDH variant rs34290285 with glioma molecular subtypes and observed significant association within IDH-mutated glioma (discovery OR = 1.54, meta P-value = 6.40 × 10⁻⁸), but not IDH wild-type glioma (meta P-value = 0.47) (Supplementary Table 12). Further stratifying IDH-mutated glioma by 1p/19q codeletion status, rs34290285 was associated with both IDH-mutated codeleted (discovery OR = 1.52, meta P-value = 0.0019) and IDH-mutated non-codeleted glioma (discovery OR = 1.64, meta P-value = 4.98 × 10⁻⁸) (Supplementary Table 12).

Discussion

Two novel glioma regions were identified to be associated with risk of specific glioma molecular subtypes in these first GWAS to be conducted by glioma molecular subtype. Variants within D2HGDH were associated with IDH-mutated glioma, and a variant near FAM20C was associated with gliomas that are IDH-mutated, TERT-mutated, and 1p/19q codeleted. We also identified a possible independent region near RTEL1; variants in GMEB2 were associated with IDH wild-type glioma. Interestingly, all three regions are located near telomeres.

We demonstrated genome-wide significance of D2HGDH variants with IDH-mutated glioma, and the more homogeneous subset of IDH-mutated non-codeleted glioma. One of these variants, rs1106639, was previously shown to have a candidate association with non-GBM (OR = 1.29, P = 1.11 × 10⁻⁵); however, it was not previously reported to be genome-wide significant.³⁰ In the meta-analyses reported herein, rs5839764 was observed to be more significant than rs1106639 in the analysis of IDH-mutated glioma. And we observed a highly significant correlation between rs71430382 (surrogate for rs5839764) genotype and D2HGDH gene expression. D2HGDH is a ubiquitously expressed enzyme found in mitochondria where it converts low levels of naturally produced D2HG to alpha-ketoglutarate (αKG).³¹ DNA interactions between rs5839764 and the 5′ region of D2HGDH in the hippocampus and cultured neural progenitor cells provide support that rs5839764 (or another variant in linkage disequilibrium) regulates expression of D2HGDH through a long-range interaction. This is further supported by the ChromHMM data that assigned promoter DNA in areas where DNA interactions were mapping to the 5′ region of D2HGDH.

A variant in D2HGDH (rs34290285) has been reported to be associated with asthma and allergic disease.^25–27 Though there have been some discrepant results in the literature, a history of allergies seems to be protective of adult diffuse glioma^28,29 and confers a better prognosis for patients who develop glioma.³² Together, these results suggest a link between the immune system and IDH mutant gliomas, mediated by D2HGDH, which may drive the glioma–allergy association. The oncometabolite 2-hydroxyglutarate is likely present at increased levels due to D2HGDH dysfunction and IDH mutation. D2HG not only promotes glioma cytosine-phosphate-guanine island methylator phenotype (G-CIMP), but also specifically plays a role in the epigenetic regulation and promotion of T-cell differentiation. Future studies are important to understand how this genetic relationship leads to altered immunobiology.³³

We observed an association with a variant near FAM20C (also known as DMP-4 and G-CK) and gliomas that have IDH mutation, TERT promoter mutation, and 1p/19q codeletion. Located near the 7p telomere, FAM20C is a promiscuous serine kinase that localizes to the lumen of the Golgi apparatus. A recent investigation of FAM20C in differentiation of human dental pulp cells found that ten-eleven translocation methylcytosine dioxygenase 1 (TET1) binds to the promoter region of FAM20C, leading to increased expression of the gene due to conversion of 5-methylcytosine (5-mc) to 5-hydroxymethylcytosine (5-hmc).³⁴FAM20C primarily phosphorylates proteins with Ser-x-Glu/pSer motifs, including approximately 80% of secreted phosphoproteins.³⁵ Known substrates of FAM20C that are implicated in various cancers include apolipoproteins,³⁶ insulin-like growth factor binding proteins,³⁷ and Serpins (serine protease inhibitors).³⁸ Changes in phosphorylation of one or more of these substrates due to altered expression of FAM20C may be responsible for the increased risk of IDH-mutated 1p/19q codeleted glioma in individuals carrying the rs111976262 variant.

We observed an association with variants in GMEB2 and IDH wild-type glioma. GMEB2 modulates glucocorticoid-mediated gene expression by binding to the glucocorticoid receptor.³⁹ The glucocorticoid receptor regulates gene expression via both transactivation of anti-inflammatory genes and transrepression of pro-inflammatory genes by binding to other transcription factors, including nuclear factor-kappaB (NF-kB).⁴⁰ Aberrant activation of the NF-kB pathway in glioma is common in IDH wild-type glioma, particularly in those classified as mesenchymal.⁴¹GMEB2 is located on chromosome arm 20q13.33, which is a region near RTEL1 that is known to be associated with adult diffuse glioma.⁸ Hi-C interactions were observed between the RTEL1 and GMEB2 regions, and each of the 2 regions appears to have DNA interactions with similar loci. This suggests that both variants may be involved in DNA-DNA interactions in cells of the brain, and that they may cooperate in the disease process.

There are some limitations with this study. The small sample sizes in some of the molecular groups limited power to detect variants with small effect sizes for these GWAS (eg, triple-negative gliomas and gliomas that are non-codeleted and have TERT and IDH mutations). Because of the limited sample size, the discovery P-value threshold was relaxed to 5 × 10⁻⁶; however, the meta-analysis P-value threshold was set at the genome-wide level of 5 × 10⁻⁸. We reported the findings of 9 GWAS studies, where each GWAS represented a particular molecular subtype. Because many of these GWAS were highly correlated, we utilized the accepted genome-wide significance threshold of 5 × 10⁻⁸. Overall, the results presented herein demonstrate that novel germline variants were detected when a GWAS is performed within more homogeneous subsets of glioma. Larger collaborative efforts across institutions will be required in order to identify variants with smaller effect sizes. It is also important to acknowledge that the experimental design utilized readily available GWAS data for both discovery and validation. The discovery set included cases and controls that were all run on the OncoArray platform, and thus imputation was performed across cases and controls. The validation set included cases that were obtained from TCGA and genotyped on the Affymetrix 6.0 genotyping array, and controls that were obtained by the Mayo Clinic Biobank and genotyped on the Illumina Omni Express genotyping array. The overlap between the Affymetrix 6.0 and Illumina Omni Express arrays was ~209 000 variants, and thus in order to achieve adequate imputation results, imputation and quality control were performed within the cases and controls separately. As an additional quality control step, the alternative allele frequencies from the discovery set for the newly identified significant variants were compared with the 1000 Genome frequencies.

The discovery of glioma germline variants has helped us to understand how gliomas arise, and has opened new avenues for etiologic research. Using 25 germline variants, patient age, and sex, glioma risk models were developed to estimate relative and lifetime absolute risks of adult diffuse glioma and subtype models to predict glioma subtypes (eg, IDH mutated vs IDH wild-type).⁴² The identification of additional novel germline variants will help to unravel the etiology of adult diffuse glioma, as well as improve the accuracy of these genetic-based risk models.

Funding

Work at Mayo Clinic was supported by National Brain Tumor Society, National Institutes of Health (P50CA108961, P30 CA15083, RC1NS068222Z, and R01 CA230712), Bernie and Edith Waterman Foundation, and Ting Tsung and Wei Fong Chao Family Foundation. Work at University of California San Francisco was supported by the National Institutes of Health (R01CA52689, P50CA097257, R01CA126831, R01CA139020, R01CA163687, and R25CA112355), as well as the Loglio Collective, the National Brain Tumor Society, the Stanley D. Lewis and Virginia S. Lewis Endowed Chair in Brain Tumor Research, the Robert Magnin Newman Endowed Chair in Neuro-oncology, and by donations from families and friends of John Berardi, Helen Glaser, Elvera Olsen, Raymond E. Cooper, and William Martinusen. Work at the University of Illinois was supported by the National Brain Tumor Society.

Conflict of interest statement. There are no conflicts of interest to disclose.

Authorship statement. Experimental design: JEEP, DHL, MW, RBJ. Implementation: TMK, TR, AC, KLD, CP, MP, HMH, LSM, PMB, LC, JW, JKW, TCB, CG. Analysis and interpretation of the data: JEEP, PAD, MLK, AMM, KLD, TMK, AB, AA, TB, JSS, SF, DHL, MW, RBJ. All authors were involved in the writing of the manuscript and have read and approved the final version.

Acknowledgments

The collection of cancer incidence data used in this study was supported by the California Department of Public Health pursuant to California Health and Safety Code Section 103885; Centers for Disease Control and Prevention’s (CDC) National Program of Cancer Registries, under cooperative agreement 5NU58DP006344; the National Cancer Institute’s Surveillance, Epidemiology and End Results Program under contract HHSN261201800032I awarded to the University of California, San Francisco, contract HHSN261201800015I awarded to the University of Southern California, and contract HHSN261201800009I awarded to the Public Health Institute, Cancer Registry of Greater California. The ideas and opinions expressed herein are those of the author(s) and do not necessarily reflect the opinions of the State of California, Department of Public Health, the National Cancer Institute, and the Centers for Disease Control and Prevention or their Contractors and Subcontractors. This publication was supported by the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health, through UCSF-CTSI Grant Number UL1 RR024131. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH. The authors acknowledge study participants, clinicians, and research staff at participating medical centers, Katherine Cornelius, Mayo Clinic Comprehensive Cancer Center Biospecimens and Processing and Genotyping Shared Resources, the UCSF Helen Diller Family Comprehensive Cancer Center Genome Analysis Core which is supported by a National Cancer Institute Cancer Center Support Grant (5P30CA082103), the UCSF Neurosurgery Tissue Bank, and The Gliogene Consortium. The results published here include data generated by the Glioma International Case Control Study established by the Gliogene Consortium and funded by NCI (R01CA139020). Information about study can be found at the following publication: The Glioma International Case-Control Study: A Report From the Genetic Epidemiology of Glioma International Consortium. Amirian ES, Armstrong GN, Zhou R, Lau CC, Claus EB, Barnholtz-Sloan JS, Il’yasova D, Schildkraut J, Ali-Osman F, Sadetzki S, Johansen C, Houlston RS, Jenkins RB, Lachance D, Olson SH, Bernstein JL, Merrell RT, Wrensch MR, Davis FG, Lai R, Shete S, Amos CI, Scheurer ME, Aldape K, Alafuzoff I, Brännström T, Broholm H, Collins P, Giannini C, Rosenblum M, Tihan T, Melin BS, Bondy ML. Am J Epidemiol. 2016 Jan 15;183(2):85–91. doi: 10.1093/aje/kwv235. The results published here are in part based upon data generated by The Cancer Genome Atlas managed by the NCI and NHGRI. Information about TCGA can be found at http://cancergenome.nih.gov.

References

Author notes