Promises, Delivery, and Challenges of Inflammatory Bowel Disease Risk Gene Discovery
2012; Elsevier BV; Volume: 11; Issue: 1 Linguagem: Inglês
10.1016/j.cgh.2012.11.001
ISSN1542-7714
Autores Tópico(s)IL-33, ST2, and ILC Pathways
ResumoOver the past two decades, investigators have used whole genome linkage and genome-wide association studies, including the "Immunochip" study, to identify a surprising number (over 163) of genetic loci containing susceptibility genes for inflammatory bowel disease (IBD) and its 2 major phenotypes, Crohn's disease (CD) and ulcerative colitis (UC). These loci, although nearly all low-risk, have provided important lessons regarding the nature of IBD etiopathogenesis, including that most loci cause both CD and UC risk; one-third of loci have risk for other common autoimmune diseases; numerous loci contain genes that regulate immunity to microbes; Th17 cells are disproportionately influenced by genes within IBD loci; and the HLA region influences UC far greater than CD. Interleukin-10 receptor subunit (IL10RA and IL10RB) and IL10 cytokine gene mutations cause a rare, severe, infantile-onset, autosomal recessive CD, and this knowledge has allowed curative treatment by bone marrow transplant. Key tasks for broader clinical translation include discovery of risk variants for non-white populations; discovery of the less frequent but higher penetrance and familial risk variants by next-generation sequencing; and determining which of numerous associated variations within loci result in specific gene dysfunction causing IBD risk—as only a small number of genes within IBD loci, including NOD2, IL23R, ATG16L1, IRGM, and PTPN22 have specific functional variations characterized. Studying the effect of IBD susceptiblity gene dysfunctions in tissue cultures and animal models will allow the ultimate translational benefits of developing novel treatments for and potentially preventing IBD in those having specific genetic risk factors. Over the past two decades, investigators have used whole genome linkage and genome-wide association studies, including the "Immunochip" study, to identify a surprising number (over 163) of genetic loci containing susceptibility genes for inflammatory bowel disease (IBD) and its 2 major phenotypes, Crohn's disease (CD) and ulcerative colitis (UC). These loci, although nearly all low-risk, have provided important lessons regarding the nature of IBD etiopathogenesis, including that most loci cause both CD and UC risk; one-third of loci have risk for other common autoimmune diseases; numerous loci contain genes that regulate immunity to microbes; Th17 cells are disproportionately influenced by genes within IBD loci; and the HLA region influences UC far greater than CD. Interleukin-10 receptor subunit (IL10RA and IL10RB) and IL10 cytokine gene mutations cause a rare, severe, infantile-onset, autosomal recessive CD, and this knowledge has allowed curative treatment by bone marrow transplant. Key tasks for broader clinical translation include discovery of risk variants for non-white populations; discovery of the less frequent but higher penetrance and familial risk variants by next-generation sequencing; and determining which of numerous associated variations within loci result in specific gene dysfunction causing IBD risk—as only a small number of genes within IBD loci, including NOD2, IL23R, ATG16L1, IRGM, and PTPN22 have specific functional variations characterized. Studying the effect of IBD susceptiblity gene dysfunctions in tissue cultures and animal models will allow the ultimate translational benefits of developing novel treatments for and potentially preventing IBD in those having specific genetic risk factors. Podcast interview: www.gastro.org/cghpodcast. Also available on iTunes. Podcast interview: www.gastro.org/cghpodcast. Also available on iTunes. Epidemiologic studies long suggested that "disease genes" (genes for which DNA variations alter disease risk) may cause inflammatory bowel disease (IBD). Crohn's disease (CD) and ulcerative colitis (UC) are increased 10-fold and 8-fold, respectively, among relatives of patients, and monozygotic twin concordance for CD (30%) and UC (15%) far exceeds dizygotic concordance (4% each), an observation best explained by underlying gene variations co-inherited by monozygotic twins that increase IBD risk.1Cho J.H. Brant S.R. Recent insights into the genetics of inflammatory bowel disease.Gastroenterology. 2011; 140: 1704-1712Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar Armed with this information, whole genome linkage (WGL) studies to identify small chromosomal segments, ie, genetic loci, that co-segregate with IBD phenotypes within multiply affected families began in the 1990s, concurrent with the development of genome-wide microsatellite markers to trace these segments. WGL studies are most useful for identifying low-frequency, higher-risk DNA variations. WGL successfully localized genes for Mendelian traits, such as cystic fibrosis, and a handful of higher penetrance genes for common, complex genetic disorders; NOD2 for CD was a notable success. The far more powerful method of genome-wide association studies (GWAS) began in IBD in 2005.1Cho J.H. Brant S.R. Recent insights into the genetics of inflammatory bowel disease.Gastroenterology. 2011; 140: 1704-1712Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar Single nucleotide polymorphisms (SNPs) (100,000 to several million) throughout the genome are tested (genotyped) on SNP platforms to determine whether a specific SNP variant, known as an allele, is associated and found significantly more frequently in unrelated cases as compared with ethnically matched controls. Because of the numerous potential independent tests throughout the genome, genome significant association is P value < 5 × 10−8; 5% false positive corrected for 1 million independent tests. GWAS has succeeded in identifying tens to more than 100 finite genetic loci for numerous phenotypes. However, the common variations tested will usually contribute only very modestly to disease risk (odds ratios [ORs], primarily 1.5 to 1.05, with few above 1.5). Moreover, even maximally associated GWAS SNPs are usually not disease causing (1) because the SNPs genotyped are only representative of SNPs in the regions and not necessarily functional and (2) because of the strong linkage disequilibrium (LD) in the human genome (Figure 1). LD means alleles of proximal SNPs have a tendency to be inherited jointly and thus are nonrandomly associated with one another. This has helped GWAS, because genotyping a given SNP (ie, a tagging SNP) can give information about numerous other proximal SNPs, especially those whose alleles are inherited tightly together in haplotype blocks. In one respect, this has been advantageous to mapping studies by reducing the number of SNPs needed to cover a given genomic region. However, LD is problematic because any one of the SNPs within a haplotype block may be the true disease-causing polymorphism (Figure 1). In addition, maximal association may be observed at non–disease-causing SNPs if functional, disease-causing alleles of multiple SNPs are in LD with a more common allele of a third SNP (which could be the case for the top association in Figure 1). Because LD is usually observed over relatively small distances, most GWAS loci may ultimately span as few as 10,000–100,000 base pairs, although such regions can contain several genes and SNPs can modify genes at far distances (eg, SNPs located within a 122,000 base pair haplotype block that map nearly 300,000 base pairs upstream from the PTGER4 gene promoter are highly associated with CD and are also highly associated with PTGER4 gene expression)1Cho J.H. Brant S.R. Recent insights into the genetics of inflammatory bowel disease.Gastroenterology. 2011; 140: 1704-1712Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar or even on other chromosomes. In both WGL and GWAS, locus identification is followed by higher density and lower allele frequency SNP genotyping to fine map (Figure 1). Ultimately, DNA sequencing is required to identify rare, unknown, and other potentially functional alleles (eg, non-SNP variants such as nucleotide insertions/deletions). Because "next-generation" sequencing costs have decreased dramatically, investigators are starting to use whole exome (coding segment) sequencing and will eventually use whole genome sequencing of cases and controls as the preferred method to directly discover disease alleles. IBD gene discovery promised to give novel insight into the underlying etiopathogenesis to determine why these enigmatic diseases occur in a hypothesis free manner; random DNA variations can be evaluated across the genome for linkage or association, independent of any purported gene functions. Although disease SNPs have been confirmed (at genome significance) with "tiny" ORs of 1.05, the discovery is meaningful because it indelibly connects the gene(s) modified to underlying IBD etiopathogenesis and may uncover pathways for therapeutic intervention. Other promises included the ability to predict (as accomplished in breast cancer with BRCA mutations years earlier) individuals at high risk for IBD, clarify heterogeneity in disease presentation and course, and ultimately identify novel targets for therapy and disease prevention. Lastly, it is important to appreciate the converse, which is that the absence of a genetic association does not preclude a role for a gene in IBD pathophysiology; genes central to IBD, such as tumor necrosis factor-α, may simply not have mutations that occurred in humans that significantly influence IBD risk. Genome linkage studies identified more than 13 IBD loci, although only IBD1, IBD3, and IBD5 have been correlated with individual, well-replicated risk genes, and only for IBD1 have specific (NOD2 R702W, G908R, and Leu1007fsX1008 loss of function [LOF]) mutations been determined as the primary source for observed linkage.1Cho J.H. Brant S.R. Recent insights into the genetics of inflammatory bowel disease.Gastroenterology. 2011; 140: 1704-1712Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar NOD2, an intracellular receptor protein in macrophages, dendritic and ileal Paneth crypt cells, activates the master immune transcription factor, nuclear factor kappa B, upon stimulation by bacterial cell wall peptidoglycans of gram-negative and gram-positive bacteria. NOD2 delivered many promises of genetics: NOD2 showed the importance of innate immunity (nonspecific immunity to microbes, usually via recognition of properties like peptidoglycans inherent to broad classes of microorganisms) to CD pathophysiology. It also showed that CD pathophysiology may be related to defective immunity to microbes because NOD2 mutations reduce or abolish the NOD2 protein's response to bacteria rather than increase it as was expected (although increased basal levels of proimmune cytokines have been observed in monocytes of CD patients with NOD2 mutations,2Brand S. Crohn's disease: Th1, Th17 or both? The change of a paradigm: new immunological and genetic insights implicate Th17 cells in the pathogenesis of Crohn's disease.Gut. 2009; 58: 1152-1167Crossref PubMed Scopus (503) Google Scholar perhaps counterbalancing the defect in immune response to bacteria). NOD2 also delivered on the promise of genetics unraveling phenotypic heterogeneity; NOD2 mutations are specific for CD vs UC and increased risk for ileal vs colon-only CD, stricturing, and internal penetrating behavior and need for intestinal resections.1Cho J.H. Brant S.R. Recent insights into the genetics of inflammatory bowel disease.Gastroenterology. 2011; 140: 1704-1712Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar In addition, NOD2 demonstrates several features later found typical for conventional (ie, non-Mendelian) IBD risk genes: (1) NOD2 risk mutations are common in the population (10% in healthy whites); (2) NOD2 contributes only slightly (1.4-fold) to familial clustering and is a risk for sporadic nearly as much as familial CD; (3) NOD2 showed the importance of ancestry in genetic risk; NOD2 LOF mutations are not found in non-white populations, with risk in African Americans proportional to the 20% European admixture3Wang M.-H. Okazaki T. Kugathasan S. et al.Contribution of higher risk genes and European admixture to Crohn's disease in African Americans.Inflamm Bowel Dis. 2012 Mar 12; ([Epub ahead of print])Crossref Scopus (26) Google Scholar; and (4) NOD2 penetrance (rate of CD among carriers) is low (0.6% mutation heterozygotes, 5% homozygotes), even though its risk far exceeds other conventional IBD risk genes (ORs, 2.4 and 17, respectively).1Cho J.H. Brant S.R. Recent insights into the genetics of inflammatory bowel disease.Gastroenterology. 2011; 140: 1704-1712Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar The other transformative linkage success came from WGL of consanguineous pedigrees with severe, infantile-onset, autosomal recessive Mendelian CD, followed by sequencing candidate genes in these regions. LOF mutations in interleukin (IL)-10 receptor subunit genes IL10RA and IL10RB and later IL-10 cytokine were identified as causes.1Cho J.H. Brant S.R. Recent insights into the genetics of inflammatory bowel disease.Gastroenterology. 2011; 140: 1704-1712Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar IL-10 and its receptors primarily function to down-regulate inflammation. Here genetics follow the conventional paradigm that CD develops from improper down-regulation of inflammation. Genetics also provided a treatment; several children homozygous for these mutations have been "cured" by allogeneic bone marrow transplanted from their heterozygote (non-CD) siblings. GWAS has identified common IL-10 genetic variations associated with a low risk (1.2-fold) of IBD.1Cho J.H. Brant S.R. Recent insights into the genetics of inflammatory bowel disease.Gastroenterology. 2011; 140: 1704-1712Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar, 4Jostins L, Ripke S, Weersma RK, et al. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature ;491:119–124.Google Scholar A CD-specific locus (OR, 1.1) was recently identified at the interferon gamma receptor 2 (IFNGR2) gene on chromosome 21, located proximal (within 110,000 base pairs) to the IL10RB gene. However, no loci for conventional IBD have been identified that overlap the IL10RA gene.4Jostins L, Ripke S, Weersma RK, et al. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature ;491:119–124.Google Scholar GWAS has been extremely productive, ultimately contributing to 163 independent loci for "conventional" IBD, 92 further confirmed and 71 newly established in the recent International IBD Genetics Consortium "Immunochip" study.4Jostins L, Ripke S, Weersma RK, et al. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature ;491:119–124.Google Scholar The Immunochip study has been the ultimate GWAS. It compiled data from 7 CD and 8 UC GWAS's in whites with replication/extension of those GWAS signals with P values < .01 performed by genotyping additional CD, UC, and control DNAs using the Immunochip genotyping chip. This chip enabled high-throughput genotyping of ∼200,000 SNPs (and a small number of insertion-deletion polymorphisms). It was designed for simultaneous replication and high-resolution fine mapping of IBD loci as well as loci for multiple other common immune-mediated disorders including ankylosing spondylitis (AS), celiac disease, multiple sclerosis, psoriasis, thyroiditis, type 1 diabetes, and systemic lupus erythematosus. For IBD, the Immunochip study evaluated combined GWAS and Immunochip genotyping data from >20,000 CD patients, >15,000 UC patients, and >25,000 controls. There are several major themes emanating from the 163 now-established IBD loci.(1)Most loci (67%) are associated with both CD and UC, with 38% showing equal risk for both phenotypes. Only 38 CD and 23 UC loci are phenotype specific (including NOD2 and IFNGR2 as noted).4Jostins L, Ripke S, Weersma RK, et al. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature ;491:119–124.Google Scholar(2)IBD shares 113 of the 163 loci with numerous traits (eg, height) and diseases, especially other immune-mediated diseases (60 of 163 loci).4Jostins L, Ripke S, Weersma RK, et al. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature ;491:119–124.Google Scholar For AS and psoriasis, 2 diseases that are found increased among IBD patients, overlap is especially remarkable (8/11 AS and 14/17 psoriasis loci).4Jostins L, Ripke S, Weersma RK, et al. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature ;491:119–124.Google Scholar When testable, the risk alleles are more often the same or in LD. An interesting exception is the functional PTPN22 Arg620Trp variant allele is associated with increased thyroiditis, rheumatoid arthritis, lupus, and type 1 diabetes risks but decreased CD risk.1Cho J.H. Brant S.R. Recent insights into the genetics of inflammatory bowel disease.Gastroenterology. 2011; 140: 1704-1712Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar(3)Numerous genes related to Th17 cell differentiation and function are associated with IBD1Cho J.H. Brant S.R. Recent insights into the genetics of inflammatory bowel disease.Gastroenterology. 2011; 140: 1704-1712Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar, 2Brand S. Crohn's disease: Th1, Th17 or both? The change of a paradigm: new immunological and genetic insights implicate Th17 cells in the pathogenesis of Crohn's disease.Gut. 2009; 58: 1152-1167Crossref PubMed Scopus (503) Google Scholar, 4Jostins L, Ripke S, Weersma RK, et al. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature ;491:119–124.Google Scholar (Figure 2). The first CD GWAS, a Japanese study, identified 2.2-fold OR alleles for CD specific to East Asian populations, noncoding variants that up-regulate expression of the TNFSF15 gene, a gene expressed in macrophages that induces Th17 differentiation.1Cho J.H. Brant S.R. Recent insights into the genetics of inflammatory bowel disease.Gastroenterology. 2011; 140: 1704-1712Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar Lower effect TNFSF15 IBD alleles were identified in whites. The second strongest signal (after NOD2) from the first GWAS in whites was a 3-fold protective Arg381Gln IL-23 receptor (IL23R) polymorphism found in 4% of CD cases and 14% of controls. At least 9 additional genes potentially influencing Th17 have been found among IBD loci1Cho J.H. Brant S.R. Recent insights into the genetics of inflammatory bowel disease.Gastroenterology. 2011; 140: 1704-1712Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar, 2Brand S. Crohn's disease: Th1, Th17 or both? The change of a paradigm: new immunological and genetic insights implicate Th17 cells in the pathogenesis of Crohn's disease.Gut. 2009; 58: 1152-1167Crossref PubMed Scopus (503) Google Scholar, 4Jostins L, Ripke S, Weersma RK, et al. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature ;491:119–124.Google Scholar (Figure 2), including IL12B that encodes the p40 subunit of both IL-12 and IL-23 cytokine heterodimers. IL12B polymorphisms have been associated with IBD, psoriasis, AS, and multiple sclerosis.1Cho J.H. Brant S.R. Recent insights into the genetics of inflammatory bowel disease.Gastroenterology. 2011; 140: 1704-1712Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar, 4Jostins L, Ripke S, Weersma RK, et al. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature ;491:119–124.Google Scholar Anti-p40 antibody, ustekinumab, was recently approved for psoriasis and is in phase 3 trials for CD.(4)In parallel to NOD2, many genes within IBD loci, especially those at greater risk for CD, influence innate immunity to microbes to the extent that genetically, CD might be a disease of dysregulated microbial immune response. A key role for autophagy dysfunction in IBD pathophysiology was highlighted by 2 gene associations. The ATG16L1 Thr300Ala variant (carried by 75% of healthy whites but 87% of those with CD) results in autophagosomes with defective ability to kill intracellular bacteria,1Cho J.H. Brant S.R. Recent insights into the genetics of inflammatory bowel disease.Gastroenterology. 2011; 140: 1704-1712Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar and IBD-associated risk polymorphisms lower IRGM expression, the IRGM protein important for induction of autophagy-induced bacterial killing. A different IRGM polymorphism is protective against tuberculosis infection.1Cho J.H. Brant S.R. Recent insights into the genetics of inflammatory bowel disease.Gastroenterology. 2011; 140: 1704-1712Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar In the IBD Immunochip study, among 8 leprosy predisposing genes (including TNFSF15), 7 map to CD loci.4Jostins L, Ripke S, Weersma RK, et al. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature ;491:119–124.Google Scholar Furthermore, multiple genes with defect-causing mutations resulting in primary immunodeficiency disorders (particularly reduced circulating T-cell or candidiasis associated) were found highly enriched within IBD loci (5-fold greater than expected, P < 10−4).4Jostins L, Ripke S, Weersma RK, et al. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature ;491:119–124.Google Scholar Six of 8 genes that cause Mendelian susceptibility to mycobacterial infection also map within IBD loci.4Jostins L, Ripke S, Weersma RK, et al. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature ;491:119–124.Google Scholar Note that the risk alleles for these diseases and IBD are usually not concordant and often have opposite functional effects.(5)HLA locus association is far more pronounced for UC than for CD. Before NOD2, the only consistent genetic associations were with HLA gene polymorphisms, especially those of the highly polymorphic HLA class II genes responsible for presentation of foreign antigens to T-lymphocytes, with greater significance for UC than for CD. In the Immunochip study, the maximal genetic association for UC (P = 5 × 10−133; OR, 1.44) at SNP rs6927022 mapped adjacent to HLA-DQA1 class II gene (between DQB1 and DRB1). There was no evidence for this SNP giving risk to CD. CD maximal association mapped to HLA class I region (P = 5 × 10−28; OR, 1.15) and showed no association with UC. These results show that the HLA region, central to autoimmunity, is especially important and broad-based (influencing a large segment of the at-risk population) for UC. For CD, the HLA association, at least in whites, appears much less important and located distally. The second strongest UC association (and also specific to UC vs CD) maps to phospholipase A2, group IIA gene (P = 2 × 10−68; OR, 1.3).4Jostins L, Ripke S, Weersma RK, et al. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature ;491:119–124.Google Scholar Most of the 163 IBD loci consist of multiple associated SNPs implicating multiple genes per locus (Figure 1). Only a handful of these loci have proven, functional, IBD-associated SNPs implicating a specific disease gene. There are 5 major tasks now. One task is GWAS and fine mapping in non-white populations. It is anticipated that just as NOD2 appears to be a European ancestral specific gene, ancestral specific loci may exist for African and Asian populations. Furthermore, the African ancestral genome in African Americans has more genetic diversity and lower LD, which can be helpful for isolating disease alleles (Figure 1). The second task is sequencing IBD cases to identify less frequent, potentially functional, higher-risk and novel mutations, including variants such as insertion-deletion polymorphisms also found commonly throughout the genome. Especially important is expansion of genome sequencing in persons with a strong IBD family history to identify the less common, higher-risk variants and potentially novel loci like IL10RA that may be disproportionately responsible for familial IBD. The third task is testing the more rare variants in cases vs controls and conditioning association of locus SNPs on the more highly associated SNPs to find less frequent but independently associated SNPs. Some genes, like IL23R, have functional coding (Arg381Glu) and independent, most likely expression-related disease variants. The fourth task is functional assays of nonconserved, amino acid–altering coding SNPs (non-synonymous SNPs). Importantly, the fifth task is determining which of numerous noncoding SNP variants affect gene expression (known as expression quantitative trait loci or "eQTLs") and RNA splicing to infer function. Ideally this should be performed by correlating genotypes of SNPs of interest with results from quantitative RNA sequencing and characterization of the effect of associated SNPs that map to epigenomically defined regulatory regions (eg, active "Enhancer" regions) in IBD relevant cells. As the true disease SNPs and the genes they modify are defined, tasks include evaluating how disease alleles interact with environmental factors and with each other. This may identify environmental triggers in those at risk and help elucidate disease pathways. Of major importance for translation will be developing materials for evaluating therapies and prevention. These include establishing cell lines and perhaps epithelial organoids from persons with higher-risk variants and establishing human disease variants (or mimics of their effect) in IBD animal models. Such studies successfully determined that norovirus can trigger Paneth cell changes similar to those seen in CD in atg16l1 low-expression mice, mimicking the effect of the 300Ala human risk variant and identifying a potential target for prevention.1Cho J.H. Brant S.R. Recent insights into the genetics of inflammatory bowel disease.Gastroenterology. 2011; 140: 1704-1712Abstract Full Text Full Text PDF PubMed Scopus (320) Google Scholar In less than 2 decades, IBD genetics has revolutionized our understanding of IBD etiopathogenesis. For those still suffering from IBD and those at risk, the outlook is steadily improving as genetic investigators complete the latter phases of disease variant discovery and increasingly focus research efforts toward determination of associated variants' functional alterations that actually result in IBD risk. The work of the genetic investigators will interconnect with other IBD investigators, including medicinal chemists and developers of biological and targeted gene therapies, immunologists, and clinicians, to develop and test specific interventions to reverse, block, or prevent these proven disease variant effects on causing IBD. Analysis shown in Figure 1 was performed by Chengrui Huang, PhD, Johns Hopkins University, using Immunochip genotyping data from the IBDGC (PIs Steven Brant; Judy Cho; Richard Duerr; Dermot McGovern; John Rioux; and Mark Silverberg with support from Mark Daly, Head of IBDGC Analytic Committee and Phil Schumm IBDGC Data Coordinating Center Manager), and additional African American control data were generously shared by Stephen Rich, University of Virginia, and Robert Kimberly, University of Alabama. /cms/asset/38b133ca-7fa5-4f7d-9caa-31d58dc13931/mmc1.mp3 Download .mp3 (10.87 MB) Help with .mp3 files Podcast
Referência(s)