While genome-wide association studies have identified genes involved in differential treatment responses to inhaled corticosteroids (ICS) in asthma, few studies have evaluated the potential effects of age in this context. A significant proportion of asthmatics experience exacerbations (hospitalizations and emergency department visits) during ICS treatment. We evaluated the interaction of genetic variation and age on ICS response (measured by the occurrence of exacerbations) through a genome-wide interaction study (GWIS) of 1,321 adult and child asthmatic patients of European ancestry. We identified 107 genome-wide suggestive (P<10−05) age-by-genotype interactions, two of which also met genome-wide significance (P<5x10-08) (rs34631960 [OR 2.3±1.6–3.3] in thrombospondin type 1 domain-containing protein 4 (THSD4) and rs2328386 [OR 0.5±0.3–0.7] in human immunodeficiency virus type I enhancer binding protein 2 (HIVEP2)) by joint analysis of GWIS results from discovery and replication populations. In addition to THSD4 and HIVEP2, age-by-genotype interactions also prioritized genes previously identified as asthma candidate genes, including DPP10, HDAC9, TBXAS1, FBXL7, and GSDMB/ORMDL3, as pharmacogenomic loci as well. This study is the first to link these genes to a pharmacogenetic trait for asthma.
Citation: Dahlin A, Sordillo JE, McGeachie M, Kelly RS, Tantisira KG, Lutz SM, et al. (2020) Genome-wide interaction study reveals age-dependent determinants of responsiveness to inhaled corticosteroids in individuals with asthma. PLoS ONE 15(3): e0229241. https://doi.org/10.1371/journal.pone.0229241
Editor: Badri Padhukasahasram, Illumina Inc, UNITED STATES
Received: September 11, 2019; Accepted: February 1, 2020; Published: March 2, 2020
Copyright: © 2020 Dahlin et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All genotype files are available from dbgap. https://www.ncbi.nlm.nih.gov/gap/.
Funding: Authors received funding from NIH (National Institutes of Health). A.C. Wu received an R01 grant (R01HD085993) and A.D. received grant K01 HL130629. Funders did not play a role in study design, analysis, data collection, decision to publish or manuscript prep. https://www.nih.gov/
Competing interests: The authors have declared that no competing interests exist.
Asthma, the most common chronic illness in childhood, costs over $50 billion annually in the United States.[1–3] Inhaled corticosteroids (ICS) are the most effective asthma controller medications leading to significant symptom improvement for most patients; however, approximately a third of individuals respond minimally or not at all. As the genetics of childhood-onset asthma appear to be distinct from adult-onset asthma, different genetic mechanisms may mediate childhood-onset versus adult-onset asthma. Furthermore, childhood-onset vs. adult-onset asthma tends to respond better to ICS treatment. In addition, national guidelines for asthma suggest distinct management approaches for childhood-onset versus adult-onset asthma.
Most studies in asthma pharmacogenetics combine data from adult and pediatric populations, despite the fact that, both clinically and phenotypically, manifestation of adult and pediatric asthma can vary widely in presentation.[8–10] Nevertheless, known associations between genes and asthma drug responses suggest that age-specific genetic mechanisms can regulate response to ICS. For example, FCER2 was demonstrated to help predict response to ICS in multiple studies of children, yet there are no reports of the role of FCER2 in ICS response in adults.[11–13] Because the distribution, number, and type of genetic polymorphisms capable of predicting asthma treatment responses may vary with changes in asthma phenotypes resulting from age, understanding how age impacts pharmacogenetic traits is important for improving treatment outcomes for patients.
The objective of this study was to identify single nucleotide polymorphisms (SNPs) that are associated with response to ICS by evaluating age-by-genotype interactions. We hypothesized that by accounting for age-by-genotype interactions, we would identify novel risk loci that predict age-specific responses to ICS in individuals with asthma.
Subjects, materials and methods
Five independent cohorts inclusive of pediatric and adult asthma patients of European ancestry were evaluated (total sample size = 1,321). The pediatric asthma population included ICS treatment arms within the Childhood Asthma Management Program (CAMP), adolescent participants from the Asthma Clinical Research Network (ACRN), and two of the five trials in the Childhood Asthma Research and Education (CARE) cohort: the Pediatric Asthma Controller Trial (PACT) and Characterizing Response to Leukotriene Receptor Antagonist and Inhaled Corticosteroid (CLIC) trials. The adult asthma cohort comprised subjects from ACRN, and data from two biorepositories linked to de-identified electronic health records from the PharmacoGenomic discovery and replication in very large POPulations (PGPop) cohorts: the Marshfield Clinic Personalized Medicine Research Project (PMRP) and Vanderbilt University Medical Center’s BioVu program (BioVu). A description of the samples from these populations used in this study is provided in S1 Table. Individuals who were present in more than one study population were removed prior to evaluation. All study procedures were approved by the respective Institutional Review Boards of each consortium and the Brigham and Women’s Hospital (the Partners Human Research Committee (PHRC)). Human Subjects approval was obtained from Partners Human Research Internal Review Board, Protocol #: 2002P000331. Written informed consent was obtained.
Phenotyping and selection of cases and controls
The main outcome for this study was a dichotomized variable for ICS response, wherein poor response (cases) was defined by one or more asthma exacerbations while on ICS and good response (controls) was defined by absence of exacerbations while on ICS. An asthma exacerbation was defined as an emergency department (ED) visit or hospitalization due to asthma, or the need for oral corticosteroids (bursts), and was assessed during the respective study period for each cohort. From a total sample size of 1,321 subjects, we selected 407 cases and 376 controls (n = 783) from CARE, ACRN, and BioVU as a discovery population, and an additional 287 cases and 251 controls (n = 538) from CAMP and PMRP for replication. We also evaluated age as an interaction variable for ICS response in GWIS models. To account for outliers as a result of the extreme influences in age ranges and right skewing in the distribution of age, we transformed age (in years) using a quantile-normalized transformation. Demographic information for cases and controls in each population is summarized in Table 1.
Genotyping, imputation, and quality control procedures
Genotyping of DNA samples from subjects enrolled in the five study populations was previously described.[14–19]. To account for differences in the genotyping arrays and platforms used for each individual study, genetic markers across all five populations were merged using PLINK v.1.94[20, 21] (https://www.cog-genomics.org/plink2), pre-phased using Shape-IT v2.,5 and imputed to the 1000 Genomes Project (phase 1 integrated release) reference CEU panels with IMPUTE2 (http://mathgen.stats.ox.ac.uk/impute/impute_v2.html). Prior to imputation, genetic markers with a minor allele frequency (MAF) < 1% and a genotype call rate < 90% were excluded, and a threshold for imputed SNPs was applied to exclude SNPs with imputation quality scores < 0.9 and MAF < 1%. Standard QC procedures were applied to the merged, imputed dataset using PLINK v.1.94 to remove samples and markers with below-threshold genotype and sample call rates, low minor allele frequency (<1%), and HWE P <10−05. Principal components analysis (PCA) was performed using PLINKv.1.94 to exclude individuals with significant non-European ancestry. A final dataset of 8,589,102 typed and imputed markers in 1,321 samples passed all sample and genotype QC measures for analysis.
Genome-wide interaction studies (GWIS) were performed in the discovery (CARE, ACRN and BioVU; n = 783) and replication (CARE and PMRP; n = 538) populations, using PLINK v.1.94. The primary analysis tested for an age-by-genotype interaction as the outcome, using logistic regression models adjusted for the main effects of age, genotype, and covariates (gender, BMI, study, and the first six principal components). To assess the significance of identified interactions, we applied statistical significance thresholds that are routinely used for genome-wide association studies. We specified a genome-wide significance threshold of 5x10-08, while we also applied a more liberal genome-wide suggestive threshold of 1x10-05 to include interactions with P-values that were slightly above genome-wide significance but that may represent genuine interactions. For the replication GWIS, interactions meeting a P-value significance threshold of 0.05 (nominal significance) were included in the joint analysis. Following the discovery GWIS, age-by-genotype interactions were filtered based on meeting both genome-wide suggestive significance (P<1x10-05) in the discovery population as well as a nominal significance threshold (P<0.05) in the replication population. For replication, the P values for any age-by-genotype interactions that met the significance thresholds for both the discovery and replication populations and which also had concordant directionality of effect estimates for the age-by-genotype interaction variable were then combined using a weighted Z-score method. Quantile-quantile plots of SNP associations demonstrated no significant inflation as a result of population stratification (S1 Fig).
Post-hoc power calculations
We conducted a post-hoc power analysis to detect age-by-genotype interactions with the binary outcome for the most significant signal, rs34631960. To run an empirical power analysis for a gene by environment interaction with a binary outcome and a normally distributed environmental exposure, we created an R package on github called powerGcE (https://github.com/SharonLutz/powerGcE). Using estimates from our study, we generated a binomial distribution for a SNP with a MAF of 0.49, the minor allele frequency of rs34631960 in our population. The transformed age variable was generated from a normal distribution with a mean of 0 and a variance of 1. Then, the binary outcome was generated using estimates from our study such that where βI varies from -1 to -0.75 by 0.05 since βI was estimated to be 0.8 in our study. We set the significance threshold at the genome-wide significance threshold of P<5x10-08. The empirical power was then calculated based on the proportion of simulations where the P-value for the interaction term in a logistic regression was less than the user specified alpha level of 5x10-08. As seen in S2 Fig, for 10,000 simulations, we had adequate power for our discovery population with 407 cases and 376 controls to detect an age by genotype interaction for rs34631960.
Our imputed datasets contained > 8 million SNPs in 1,321 pediatric and adult patients with asthma who were taking ICS. We evaluated age-by-genotype interactions with the outcome of ICS response in two independent (discovery and replication) GWIS. The discovery GWIS of 407 asthmatic cases with poor ICS response and 376 asthmatic controls with good ICS response (total n = 783; Table 1) identified 230 age-by-genotype interactions that were suggestive of genome-wide significance (P<1x10-05) (Table 2); however, none achieved genome-wide significance (P<5x10-08) (Fig 1). Among the suggestive associations, 107 SNPs were also at least nominally significant in the replication population. Of these, two SNPs (rs34631960 and rs2328386) achieved genome-wide significance by joint analysis (Table 2). Sixty-nine of the 107 replicated GWIS SNPs were localized to intergenic regions of chromosomes 2, 5, 11, and 12, while the remainder were distributed across eight unique genes in chromosomes 1 (SPRR2G and SAMD13), 6 (HIVEP2), 8 (SAMD12), 15 (THSD4), and 16 (RBFOX1). All but one of the genic SNPs were predicted to reside within the introns of their respective genes, while rs509194 was annotated to a 3’UTR in SPRR2G.
Manhattan plots show results for discovery (top panel) and replication (bottom panel) GWIS, by chromosome. Red line indicates threshold for genome-wide significance (P<5x10-08) while blue lines indicate suggestive threshold for genome-wide significance (P<10−05).
The top two SNPs, rs34631960 and rs2328386, were localized to thrombospondin type 1 domain-containing protein 4 (THSD4) and human immunodeficiency virus type I enhancer binding protein 2 (HIVEP2), respectively (Table 2). rs34631960 was associated with an approximately 2-fold increase in risk of exacerbations on ICS with age, while rs2328386 was associated with a 0.3–0.5-fold decrease in risk of exacerbations on ICS with age (Table 2).
We also investigated age-by-genotype interactions for candidate genes within inflammatory pathways which were previously identified in studies of ICS response or asthma susceptibility. These genes included dipeptidyl-peptidase 10 (DPP10), histone deacetylase 9 (HDAC9), thromboxane A synthase 1 (TBXAS1), F-box and leucine rich repeat protein 7 (FBXL7), and gasdermin B (GSDMB/ORMDL sphingolipid biosynthesis regulator 3 (ORMDL3) locus. All genes had at least nominally significant SNP annotations in both the discovery and replication results; when ranking these candidate gene variants by the lowest P-values, DPP10 was the top gene in the discovery set, while HDAC9 was the top gene in the replication set (S2 Table). The top SNPs in both discovery and replication sets for DPP10 were associated with lower risk of exacerbations on ICS with age, which was also observed for the top SNPs in HDAC9 for both sets (S2 Table).
Our study demonstrates three key findings. First, the age-by-genotype interaction analysis identified two SNPs, rs34631960 in THSD4 and rs2328386 in HIVEP2, that met genome-wide significance based on meta-analysis in 1,321 pediatric and adult patients with asthma who were taking ICS. To our knowledge, this is the first genome-wide study to link THSD4 and HIVEP2 to a pharmacogenomic outcome for asthma. Secondly, we identified additional novel age-by-genotype interactions related to ICS response. Lastly, our study underscores the importance of investigating age-by-genotype interactions in asthma.
The top-ranked age-by-genotype association (THSD4 SNP rs34631960) could confer a protective effect on exacerbations risk in younger asthmatics taking ICS, or, conversely, may predict an increased risk of poor ICS response with increasing age. THSD4 is potentially involved in regulating normal and abnormal lung function, angiogenesis, and airway remodeling in asthma and may also contribute to the progression to asthma, asthma severity, and exacerbations through these processes.[26–28] THSD4 was also previously associated with COPD, and also with measures of lung function in individuals with and without asthma, including FEV1/FVC ratio in large meta-analyses of individuals of European ancestry.[30–33] Consistent with these findings, our results demonstrate the association of THSD4 with asthma exacerbations and further suggest a role for variation within THSD4 as a potential mecahnism for increased risk of exacerbations with age, despite ICS use. There is little evidence to suggest that THSD4 may be regulated by corticosteroid exposure. However, we observed that THSD4 was differentially expressed in response to dexamethasone treatment of lymphoblastoid cell lines (LCLs) from individuals with asthma, although the [log2]abs fold change in expression was below threshold (data not shown). Additional studies are warranted to discern the mechanistic role(s) of THSD4 in ICS response.
rs2328386 confers a C>T change in the large, zinc-finger containing transcription factor, HIVEP2. Like rs34631960, rs2328386 was also annotated to an intronic region; therefore, the SNP may be unlikely to have direct functional effects. HIVEP2, also known as Schnurri (Shn)-2, contributes to asthma pathogenesis through its role as an NFkB binding protein, and regulates Th2 cell differentiation and responses. Furthermore, HIVEP2 is also a negative regulator of Th2-mediated allergic airway inflammation; Shn-2 deficient mice demonstrated increased eosinophilic inflammation, mucus hyperproduction and airway hyperresponsiveness. HIVEP2 is likely to be involved in other asthma phenotypes, as it mediates multiple aspects of growth and development for a variety of biological pathways and tissues, notably the nervous system, immune response pathways,[36–38] and bone remodeling. HIVEP2 could potentially contribute to asthma severity and drug response through glucocorticoid receptor activity; Hivep2 expression is inducible in dexamethasone-stimulated, mouse bone-marrow-derived macrophages, via glucocorticoid receptor-mediated effects on chromatin reorganization. Furthermore, as it also binds to the enhancers of multiple genes in the MHC region of chromosome 6, HIVEP2 regulates the activity of a variety of additional genes involved in immune responses, inflammation and glucocorticoid response.[41,42] Variation in HIVEP2 has also been noted as a contributing factor for neurodevelopmental disorders, such as schizophrenia, and HIVEP2 is thought to promote the activation of inflammatory pathways precipitating these diseases. Our findings suggest that HIVEP2 variation may also contribute to age-dependent corticosteroid response in asthma patients.
To date, multiple genes that influence allergic inflammatory pathways have been associated with asthma phenotypes including DPP10, which is implicated in asthma susceptibility,[44, 45] HDAC9, which is associated with asthma severity, TBXAS1 and FBXL7, which are associated with ICS response in children, and the GSDMB/ORMDL3 locus, which is associated with child-onset asthma and is also influential for ICS response. Due to their importance in inflammatory pathways related to asthma phenotypes, we tested for association of these genes in the age-by-genotype interaction results using the top-ranked SNP within 10 kb of each candidate gene as a proxy, and identified DPP10 and HDAC9 as the most significantly associated genes for the discovery and replication datasets. The top SNPs in both genes were associated with lower risk of exacerbations while taking ICS, in both cohorts (S2 Table). While DPP10 and HDAC9 were implicated as risk loci for asthma susceptibility and/or severity and related phenotypes, this study is the first to indicate an association of these genes with ICS response. Furthermore, we confirmed previously reported associations of TBXAS1, FBXL7, and GSDMB/ORMDL3 with ICS response, and also contribute several new age-by-genotype interactions to this list.
Strengths of this study comprise the inclusion of multiple clinical trials and ‘real-life’ asthma populations of diverse age ranges, spanning childhood, adolescence, and adulthood. To our knowledge, no previous pharmacogenetic studies have specifically examined the effect of age-by-genotype interactions on the outcome of differential ICS responsiveness, as measured by the occurrence of exacerbations concurrent with ICS use. However, we acknowledge that there are limitations with the selection of exacerbations as a pharmacogenomic phenotype for ICS response, because in addition to patient genetics, other important factors may also contribute to exacerbations on ICS, such as medication adherence, environmental conditions, and disease-related co-morbidities. The cohorts investigated in this study were derived from well-characterized asthma clinical trials and biorepositories, and we excluded patients with respiratory-related comorbidities that might influence exacerbations, such as COPD, bronchiectasis, and pulmonary hypertension. Combining heterogeneous data from different clinical study and patient populations is further hindered by the fact that additional relevant information including ICS dose, atopy status, etc., is not universally available across studies. While it was difficult to ascertain the extent of ICS adherence or dose (exposure) in these populations, as this information is not readily available across cohorts, to avoid including subjects with poor adherence, our study populations were restricted to patients with an EMR of concurrent ICS prescription throughout the two-year study window. We attempted to minimize bias introduced by combining data from multiple studies, through standardizing the definition of exacerbations across studies and by merging genotype data (which was originally generated using different array platforms). Further, while we sought to capture ages of the cases reported at the time of the exacerbations event, in the absence of this information, we analyzed age as recorded at the two-year study visit instead, which was appropriate given that exacerbations were evaluated cumulatively through this time point. Despite this, we acknowledge that differences across studies and populations could represent a limitation. In addition, our study was restricted to subjects of European ancestry due to the larger sample sizes available for investigation, which could limit generalizability of our findings to the asthma patient population. As a result, future age-by-genotype analyses in diverse populations are warranted. While the causes of exacerbations are multi-factorial, our study design is tailored to capture specific genetic associations with the ICS response outcome as modified by age, in patients who are taking ICS. However, as we evaluated only ICS treated asthma patients (vs. a non-ICS treated group), we acknowledge that it is difficult to discern if the interactions we identified reflect those that differ by age in both ICS treated and untreated populations. Further, our sample sizes may not be large enough to detect certain interactions of lower significance levels that could potentially contribute to the phenotype; however, the sample size of our population is relatively large for pharmacogenomics GWAS, particularly of asthma, and post-hoc power analyses showed that our study was adequately powered (S2 Fig). While we did not identify genome-wide significant associations in the discovery analysis, we were able to identify associations that reached genome-wide significance through joint analysis of both discovery and replication populations. As the top age-by-genotype interactions were annotated to introns of THSD4 and HIVEP2, they are unlikely to confer direct functional effects with regard to the phenotype. Furthermore, neither SNP is an eQTL in the Genotype-Tissue Expression (GTEx) portal (https://gtexportal.org/home/) for any gene or tissue. Despite this, the present study, in addition to previous literature defining roles for both THSD4 and HIVEP2 in asthma, show that these may represent candidate genes for which additional molecular studies are warranted in order to clarify their roles in asthma and differential ICS response.
In summary, we have identified novel, age-dependent genetic polymorphisms associated with ICS response in adult and pediatric asthma patients. We report that THSD4, HIVEP2, DPP10, HDAC9 and other genes within inflammatory response pathways and with definitive roles in asthma susceptibility, severity, and exacerbations, are also asthma pharmacogenes. These findings contribute to precision medicine-driven efforts to surmount treatment-resistant asthma in patients of all ages.
S1 Fig. Quantile-quantile plots of GWIS for (A) discovery and (B) replication.
S1 Table. Summary of clinical trials and patient populations.
- 1. National Institute of Allergy and Infectious Disease (2001). Asthma: A Concern for Minority Populations. NIAID, NIH.
- 2. (2001). Condition of Education. NCES, Department of Education.
- 3. American Lung Association (2012). Asthma in Adults Fact Sheet.
- 4. Chan MT, Leung DY, Szefler SJ, Spahn JD. Difficult-to-control asthma: clinical characteristics of steroid-insensitive asthma. J Allergy Clin Immunol 1998; 101(5): 594–601. pmid:9600494
- 5. Moffatt MF, Gut IG, Demenais F, Strachan DP, Bouzigon E, Heath S, et al. A large-scale, consortium-based genomewide association study of asthma. N Engl J Med 2010; 363(13): 1211–1221. pmid:20860503
- 6. Kaditis AG, Winnie G, Syrogiannopoulos GA. Anti-inflammatory pharmacotherapy for wheezing in preschool children. Pediatr Pulmonol 2007; 42(5): 407–420. pmid:17358042
- 7. Expert Panel Report 3 (EPR-3): Guidelines for the Diagnosis and Management of Asthma-Summary Report 2007. J Allergy Clin Immunol 2007; 120(5 Suppl): S94–138. pmid:17983880
- 8. Litonjua AA, Lasky-Su J, Schneiter K, Tantisira KG, Lazarus R, Klanderman B, et al. ARG1 is a novel bronchodilator response gene: screening and replication in four asthma cohorts. Am J Respir Crit Care Med 2008; 178(7): 688–694. pmid:18617639
- 9. Himes BE, Jiang X, Hu R, Wu AC, Lasky-Su JA, Klanderman BJ, et al. Genome-wide association analysis in asthma subjects identifies SPATS2L as a novel bronchodilator response gene. PLoS Genet 2012; 8(7): e1002824. pmid:22792082
- 10. Wu AC, Tantisira K, Li L, Schuemann B, Weiss S. Repeatability of response to asthma medications. J Allergy Clin Immunol 2009; 123(2): 385–390. pmid:19064281
- 11. Tantisira KG, Silverman ES, Mariani TJ, Xu J, Richter BG, Klanderman BJ, et al. FCER2: a pharmacogenetic basis for severe exacerbations in children with asthma. J Allergy Clin Immunol 2007; 120(6): 1285–1291. pmid:17980418
- 12. Koster ES, Maitland-van der Zee AH, Tavendale R, Mukhopadhyay S, Vijverberg SJ, Raaijmakers JA, et al. FCER2 T2206C variant associated with chronic symptoms and exacerbations in steroid-treated asthmatic children. Allergy 2011; 66(12): 1546–1552. pmid:21958076
- 13. Szalai R, Matyas P, Varszegi D, Melegh M, Magyari L, Jaromi L, et al. Admixture of beneficial and unfavourable variants of GLCCI1 and FCER2 in Roma samples can implicate different clinical response to corticosteroids. Molecular biology reports 2014; 41(11): 7665–7669. pmid:25091946
- 14. The Childhood Asthma Management Program (CAMP): design, rationale, and methods. Childhood Asthma Management Program Research Group. Control Clin Trials 1999; 20(1): 91–120. pmid:10027502
- 15. Boushey HA, Sorkness CA, King TS, Sullivan SD, Fahy JV, Lazarus SC, et al. Daily versus as-needed corticosteroids for mild persistent asthma. N Engl J Med 2005; 352(15): 1519–1528. pmid:15829533
- 16. Sorkness CA, Lemanske RF, Mauger DT, Boehmer SJ, Chinchilli VM, Martinez FD, et al. Long-term comparison of 3 controller regimens for mild-moderate persistent childhood asthma: the Pediatric Asthma Controller Trial. J Allergy Clin Immunol 2007; 119(1): 64–72. pmid:17140647
- 17. Szefler SJ, Phillips BR, Martinez FD, Chinchilli VM, Lemanske RF, Strunk RC, et al. Characterization of within-subject responses to fluticasone and montelukast in childhood asthma. J Allergy Clin Immunol 2005; 115(2): 233–242. pmid:15696076
- 18. McCarty CA, Wilke RA, Giampietro PF, Wesbrook SD, Caldwell MD. Marshfield Clinic Personalized Medicine Research Project (PMRP): design, methods and recruitment for a large population-based biobank. Per Med 2005; 2(1): 49–79. pmid:29793241
- 19. Roden DM, Pulley JM, Basford MA, Bernard GR, Clayton EW, Balser JR, et al. Development of a large-scale de-identified DNA biobank to enable personalized medicine. Clin Pharmacol Ther 2008; 84(3): 362–369. pmid:18500243
- 20. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 2007; 81(3): 559–575. pmid:17701901
- 21. Chang CC, Chow CC, Tellier LC, Vattikuti S, Purcell SM, Lee JJ. Second-generation PLINK: rising to the challenge of larger and richer datasets. Gigascience 2015; 4: 7. pmid:25722852
- 22. Delaneau O, Coulonges C, Zagury JF. Shape-IT: new rapid and accurate algorithm for haplotype inference. BMC Bioinformatics 2008; 9: 540. pmid:19087329
- 23. Kvale MN, Hesselson S, Hoffmann TJ, Cao Y, Chan D, Connell S, et al. Genotyping Informatics and Quality Control for 100,000 Subjects in the Genetic Epidemiology Research on Adult Health and Aging (GERA) Cohort. Genetics 2015; 200(4): 1051–1060. pmid:26092718
- 24. Howie BN, Donnelly P, Marchini J. A flexible and accurate genotype imputation method for the next generation of genome-wide association studies. PLoS Genet 2009; 5(6): e1000529. pmid:19543373
- 25. Liptak T. On the combination of independent tests. Magyar Tud Akad Mat Kutato Int Kozl 1958; 3: 171–197.
- 26. Chen H, Herndon ME, Lawler J. The cell biology of thrombospondin-1. Matrix Biol 2000; 19(7): 597–614. pmid:11102749
- 27. Favre CJ, Mancuso M, Maas K, McLean JW, Baluk P, McDonald DM. Expression of genes involved in vascular development and angiogenesis in endothelial cells of adult lung. Am J Physiol Heart Circ Physiol 2003; 285(5): H1917–1938. pmid:12842817
- 28. Yao TC, Du G, Han L, Sun Y, Hu D, Yang JJ, et al. Genome-wide association study of lung function phenotypes in a founder population. J Allergy Clin Immunol 2014; 133(1): 248–255.e241–210. pmid:23932459
- 29. Morrow JD, Cho MH, Hersh CP, Pinto-Plata V, Celli B, Marchetti N, et al. DNA methylation profiling in human lung tissue identifies genes associated with COPD. Epigenetics 2016: 1–10.
- 30. Soler Artigas M, Loth DW, Wain LV, Gharib SA, Obeidat M, Tang W, et al. Genome-wide association and large-scale follow up identifies 16 new loci influencing lung function. Nat Genet 2011; 43(11): 1082–1090. pmid:21946350
- 31. Hancock DB, Eijgelsheim M, Wilk JB, Gharib SA, Loehr LR, Marciante KD, et al. Meta-analyses of genome-wide association studies identify multiple loci associated with pulmonary function. Nat Genet 2010; 42(1): 45–52. pmid:20010835
- 32. Repapi E, Sayers I, Wain LV, Burton PR, Johnson T, Obeidat M, et al. Genome-wide association study identifies five loci associated with lung function. Nat Genet 2010; 42(1): 36–44. pmid:20010834
- 33. Li X, Hawkins GA, Ampleford EJ, Moore WC, Li H, Hastie AT, et al. Genome-wide association study identifies TH1 pathway genes associated with lung function in asthmatic patients. J Allergy Clin Immunol 2013; 132(2): 313–320.e315. pmid:23541324
- 34. Iwamura C, Kimura MY, Shinoda K, Endo Y, Hasegawa A, Yamashita M, et al. Schnurri-2 regulates Th2-dependent airway inflammation and airway hyperresponsiveness. Int Immunol 2007; 19(6): 755–762. pmid:17493960
- 35. Srivastava S, Engels H, Schanze I, Cremer K, Wieland T, Menzel M, et al. Loss-of-function variants in HIVEP2 are a cause of intellectual disability. Eur J Hum Genet 2016; 24(4): 556–561. pmid:26153216
- 36. Yamashita J, Iwamura C, Mitsumori K, Hosokawa H, Sasaki T, Takahashi M, et al. Murine Schnurri-2 controls natural killer cell function and lymphoma development. Leuk Lymphoma 2012; 53(3): 479–486. pmid:21936769
- 37. Staton TL, Lazarevic V, Jones DC, Lanser AJ, Takagi T, Ishii S, et al. Dampening of death pathways by schnurri-2 is essential for T-cell development. Nature 2011; 472(7341): 105–109. pmid:21475200
- 38. Nakayama T, Kimura MY. Memory Th1/Th2 cell generation controlled by Schnurri-2. Adv Exp Med Biol 2010; 684: 1–10. pmid:20795536
- 39. Saita Y, Takagi T, Kitahara K, Usui M, Miyazono K, Ezura Y, et al. Lack of Schnurri-2 expression associates with reduced bone remodeling and osteopenia. J Biol Chem 2007; 282(17): 12907–12915. pmid:17311925
- 40. Jubb AW, Boyle S, Hume DA, Bickmore WA. Glucocorticoid Receptor Binding Induces Rapid and Prolonged Large-Scale Chromatin Decompaction at Multiple Target Loci. Cell Rep 2017; 21(11): 3022–3031. pmid:29241532
- 41. Fukuda S, Yamasaki Y, Iwaki T, Kawasaki H, Akieda S, Fukuchi N, et al. Characterization of the biological functions of a transcription factor, c-myc intron binding protein 1 (MIBP1). J Biochem 2002; 131(3): 349–357. pmid:11872163
- 42. Kehoe RF, Zheutlin TA, Dunnington CS, Mattioni TA, Yu G, Spangenberg RB. Safety and efficacy of sotalol in patients with drug-refractory sustained ventricular tachyarrhythmias. Am J Cardiol 1990; 65(2): 58A–64A; discussion 65A-66A. pmid:2294689
- 43. Takao K, Kobayashi K, Hagihara H, Ohira K, Shoji H, Hattori S, et al. Deficiency of schnurri-2, an MHC enhancer binding protein, induces mild chronic inflammation in the brain and confers molecular, neuronal, and behavioral phenotypes related to schizophrenia. Neuropsychopharmacology 2013; 38(8): 1409–1425. pmid:23389689
- 44. Torgerson DG, Capurso D, Mathias RA, Graves PE, Hernandez RD, Beaty TH, et al. Resequencing candidate genes implicates rare variants in asthma susceptibility. Am J Hum Genet 2012; 90(2): 273–281. pmid:22325360
- 45. Agrawal DK, Shao Z. Pathogenesis of allergic airway inflammation. Curr Allergy Asthma Rep 2010; 10(1): 39–48. pmid:20425513
- 46. Hou X, Wan H, Ai X, Shi Y, Ni Y, Tang W, et al. Histone deacetylase inhibitor regulates the balance of Th17/Treg in allergic asthma. Clin Respir J 2016; 10(3): 371–379. pmid:25307458
- 47. Park HW, Dahlin A, Tse S, Duan QL, Schuemann B, Martinez FD, et al. Genetic predictors associated with improvement of asthma symptoms in response to inhaled corticosteroids. J Allergy Clin Immunol 2014; 133(3): 664–669 e665. pmid:24486069
- 48. Berce V, Kozmus CE, Potocnik U. Association among ORMDL3 gene expression, 17q21 polymorphism and response to treatment with inhaled corticosteroids in children with asthma. The pharmacogenomics journal 2013; 13(6): 523–529. pmid:22986918