Family-based exome-wide association study of childhood acute lymphoblastic leukemia among Hispanics confirms role of ARID5B in susceptibility

Abstract

We conducted an exome-wide association study of childhood acute lymphoblastic leukemia (ALL) among Hispanics to confirm and identify novel variants associated with disease risk in this population. We used a case-parent trio study design; unlike more commonly used case-control studies, this study design is ideal for avoiding issues with population stratification bias among this at-risk ethnic group. Using 710 individuals from 323 Guatemalan and US Hispanic families, two inherited SNPs in ARID5B reached genome-wide level significance: rs10821936, RR = 2.31, 95% CI = 1.70–3.14, p = 1.7×10⁻⁸ and rs7089424, RR = 2.22, 95% CI = 1.64–3.01, p = 5.2×10⁻⁸. Similar results were observed when restricting our analyses to those with the B-ALL subtype: ARID5B rs10821936 RR = 2.22, 95% CI = 1.63–3.02, p = 9.63×10⁻⁸ and ARID5B rs7089424 RR = 2.13, 95% CI = 1.57–2.88, p = 2.81×10⁻⁷. Notably, effect sizes observed for rs7089424 and rs10821936 in our study were >20% higher than those reported among non-Hispanic white populations in previous genetic association studies. Our results confirmed the role of ARID5B in childhood ALL susceptibility among Hispanics; however, our assessment did not reveal any strong novel inherited genetic risks for acute lymphoblastic leukemia among this ethnic group.

About the Authors:

Natalie P. Archer

Contributed equally to this work with: Natalie P. Archer, Virginia Perez-Andreu

Roles Formal analysis, Methodology, Writing – original draft

Affiliation: University of Texas School of Public Health, Austin Regional Campus, Austin, Texas, United States of America

Virginia Perez-Andreu

Contributed equally to this work with: Natalie P. Archer, Virginia Perez-Andreu

Roles Formal analysis, Writing – original draft

Affiliations Department of Pharmaceutical Sciences, St. Jude Children's Research Hospital, Memphis, Tennessee, United States of America, Hematologic Malignancies Program, Comprehensive Cancer Center, St. Jude Children’s Research Hospital, Memphis, Tennessee, United States of America

Ulrik Stoltze

Roles Formal analysis

Affiliation: University Hospital Copenhagen, Copenhagen, Denmark

Michael E. Scheurer

Roles Methodology, Supervision

Affiliation: Department of Pediatrics, Section of Hematology-Oncology, Baylor College of Medicine, Houston, Texas, United States of America

Anna V. Wilkinson

Roles Supervision

Affiliation: Michael & Susan Dell Center for Healthy Living, University of Texas Health Science Center at Houston, School of Public Health, Austin Regional Campus, Austin, Texas, United States of America

Ting-Nien Lin

Roles Resources

Affiliation: Department of Pharmaceutical Sciences, St. Jude Children's Research Hospital, Memphis, Tennessee, United States of America

Maoxiang Qian

Roles Resources

Affiliation: Department of Pharmaceutical Sciences, St. Jude Children's Research Hospital, Memphis, Tennessee, United States of America

Charnise Goodings

Roles Resources

Affiliations Department of Pharmaceutical Sciences, St. Jude Children's Research Hospital, Memphis, Tennessee, United States of America, Hematologic Malignancies Program, Comprehensive Cancer Center, St. Jude Children’s Research Hospital, Memphis, Tennessee, United States of America

Michael D. Swartz

Roles Supervision

Affiliation: Department of Biostatistics, School of Public Health, The University of Texas Health Science Center at Houston, Houston, Texas, United States of America

Nalini Ranjit

Roles Supervision

Affiliation: Michael & Susan Dell Center for Healthy Living, University of Texas Health Science Center at Houston, School of Public Health, Austin Regional Campus, Austin, Texas, United States of America

Karen R. Rabin

Roles Supervision

Affiliation: Department of Pediatrics, Section of Hematology-Oncology, Baylor College of Medicine, Houston, Texas, United States of America

Erin C. Peckham-Gregory

Roles Resources

Affiliation: Department of Pediatrics, Section of Hematology-Oncology, Baylor College of Medicine, Houston, Texas, United States of America

Sharon E. Plon

Roles Funding acquisition

Affiliation: Department of Pediatrics, Section of Hematology-Oncology, Baylor College of Medicine, Houston, Texas, United States of America

Pedro A. de Alarcon

Roles Data curation

Affiliation: Department of Pediatrics, University of Illinois College of Medicine at Peoria, Peoria, Illinois, United States of America

Ryan C. Zabriskie

Roles Data curation

Affiliation: Department of Pediatrics, Section of Hematology-Oncology, Baylor College of Medicine, Houston, Texas, United States of America

Federico Antillon-Klussmann

Roles Data curation

Affiliations Unidad Nacional de Oncología Pediátrica, Guatemala City, Guatemala, School of Medicine, Francisco Marroquín University, Guatemala City, Guatemala

Cesar R. Najera

Roles Data curation

Affiliation: Unidad Nacional de Oncología Pediátrica, Guatemala City, Guatemala

Jun J. Yang

Roles Conceptualization, Funding acquisition, Supervision, Writing – review & editing

* E-mail: [email protected] (J.J.Y.); [email protected] (P.J.L.)

Affiliations Department of Pharmaceutical Sciences, St. Jude Children's Research Hospital, Memphis, Tennessee, United States of America, Hematologic Malignancies Program, Comprehensive Cancer Center, St. Jude Children’s Research Hospital, Memphis, Tennessee, United States of America

Philip J. Lupo

Roles Conceptualization, Formal analysis, Funding acquisition, Methodology, Supervision, Writing – review & editing

* E-mail: [email protected] (J.J.Y.); [email protected] (P.J.L.)

Affiliation: Department of Pediatrics, Section of Hematology-Oncology, Baylor College of Medicine, Houston, Texas, United States of America

ORCID http://orcid.org/0000-0003-0978-5863

Introduction

Acute lymphoblastic leukemia (ALL) is the most common malignancy among children, with B-cell ALL (B-ALL) accounting for the majority (80% to 85%) of cases [1,2]. Genome-wide association studies (GWAS) have identified several inherited genetic variants associated with childhood or adolescent ALL risk, including but not limited to single nucleotide polymorphisms (SNPs) in ARID5B, IKZF1, CEBPE, CDKN2A, PIP4K2A, GATA3, LHPP, and ELK3 [3–11]. Another important risk factor for childhood ALL is Hispanic ethnicity. Children of Hispanic ethnic background have a 10% to 30% higher incidence of ALL than do non-Hispanic whites, and a rate almost two times higher than among non-Hispanic blacks [12,13]. Hispanic children with ALL also have a lower 5-year survival rate and a higher incidence of relapse than do non-Hispanic whites [14,15].

These differences in ALL incidence and outcomes among Hispanics could be due to differences in the frequency of known or novel genetic risk factors that are unique to this population. Genetic risk factors associated with Amerindian ancestry could account for increased ALL incidence and decreased ALL survival among Hispanics [15]. Furthermore, risk alleles in ARID5B and GATA3 were found to be more frequent among Hispanic Americans than in European Americans [8,9], consistent with ancestry-related disparities in ALL susceptibility and treatment outcomes. However, there is much work remaining to identify the missing heritability of ALL among Hispanics.

While a majority of GWAS rely on the case-control study design, this approach is subject to population stratification bias. This bias may result in spurious associations or the masking of associations due to subgroups within a population that have different genetic profiles and/or frequencies of disease [16,17]. Population stratification bias can be particularly important among Hispanic populations, where a three-way admixture among Native American, European, and West African ancestry populations is common [18,19]. However, the family-based case-parent trio design is immune to population stratification bias, as this approach relies on evaluating disequilibrium in the transmission of alleles among affected cases and their parents, which does not vary by race and ethnicity [16,20]. Thus, the case-parent trio study design has an advantage over the case-control study when analyzing data from highly admixed populations such as Hispanics.

To determine if previous ALL GWAS findings in populations primarily of European descent are transferable to Hispanic populations, and to identify novel genetic risks for ALL for this ethnic group, we conducted an exome-wide association study (EXWAS) among Hispanics recruited in Guatemala and the Southwestern United States. We used a case-parent trio approach to attenuate the potential for population stratification bias due to admixture among this population.

Materials and methods

Study subjects and samples

The study population has been described previously [21]. Briefly, the population under study included Guatemalan ALL case-parent trios recruited during the period 2012–2015 from Unidad Nacional de Oncología Pediátrica (UNOP), a pediatric cancer treatment center in Guatemala City, as well as Hispanic trios recruited during the period 2003–2011 from the Texas Children’s Cancer Center (part of Texas Children’s Hospital (TCH) in Houston, Texas). Children and adolescents (ages 1 to 19 years) diagnosed with ALL and treated at either UNOP or TCH during their respective recruitment timeframes were eligible for the study. All Guatemalan ALL cases were either of indigenous Native American ancestry (Mayan; primarily K’iche’, Mam, and Q’anjob’al) or Ladino (Mestizo, with a mixed Native American and European background) ancestry. Written informed consent was obtained from parents, guardians, or patients, as appropriate. For all ALL cases, germline genomic DNA was extracted from peripheral blood samples obtained during clinical remission. Peripheral blood samples were also obtained from participating Guatemalan parents, while saliva samples were collected from parents of TCH cases. DNA extraction was performed using a Puregene DNA isolation kit by Gentra Systems.

This study was approved by the Institutional Review Boards at University of Texas Health Science Center (UTHSC), Baylor College of Medicine, and St. Jude Children’s Research Hospital, as well as by the Bioethics Committee at Facultad de Medicina, Universidad Francisco Marroquín in Guatemala.

Genotyping and quality control (QC)

Germline genomic DNA for all samples was submitted to exome-wide genotyping using the Illumina Human Exome BeadChip (Illumina, San Diego CA). SNPs genotyped (N = 237,436) were mostly from exonic regions. Genotype calls were made using Illumina GenomeStudio Software; genotype was coded as 0, 1, or 2, indicating the number of minor alleles (an additive genetic model was assumed). Individuals with a genotype call rate <95% and case-parent trios or duos with a Mendelian error rate (i.e., Mendelian inconsistencies) >0.5% were excluded from analyses. In addition, non-autosomal SNPs, SNPs with a minor allele frequency (MAF) <1%, and SNPs with poor genotyping quality (<95% call rate) were excluded. For SNPs with a MAF of 1%-5%, a more stringent set of QC criteria determined by call rate was used (S1 Fig), and SNPs that did not meet these criteria were also excluded from analyses.

Statistical analysis

For each SNP, associations between the inherited genotype and risk of childhood ALL were assessed with multinomial modeling, using the EMIM program [22]. This analysis compares observed case genotype distributions to expected distributions, assuming Mendelian transmission of the minor allele [20,22,23]. Multinomial modeling is mathematically equivalent to a log-linear approach [24], which has been used in several case-parent trio GWAS [25–27]. The multinomial modeling approach allows for the direct estimation of inherited effects even when one or more individuals are missing from a case-parent trio [24]. Case-parent trios, case-mother or case-father duos, parents alone, or cases alone can all be analyzed simultaneously in EMIM to estimate inherited genetic effects.

We used EMIM to estimate relative risks (RR) for each SNP, as well as corresponding 95% confidence intervals (CIs) and chi-squared values, assuming a log-additive model of inheritance. EMIM analyses stratified the data by parental mating type, inherently adjusting RR estimates for effects of population stratification [23,24]. We calculated two-tailed p-values from chi-squared values for each SNP using R, version 3.2.2. We also created Manhattan plots and Q-Q plots using R; and regional association plots using LocusZoom [28]. We used a threshold of P<1.0 × 10⁻⁶ to denote statistical significance in our analyses. The majority of ALL cases in the study cohort were of the B-ALL subtype; therefore, in addition to examining inherited variants associated with ALL in the entire study cohort, we also conducted a subgroup analysis with only B-ALL cases and families, to assess inherited genotype effects for B-ALL in particular.

In order to select cases for the TCH population, we determined genetic ancestry using STRUCTURE [15,29] on the basis of genotypes at 30,000 randomly selected SNPs, using HapMap samples (CEU, YRI, CHB/JPT) and indigenous Amerindian references [30] as ancestry populations. Hispanic Americans were defined as individuals for whom the proportion of Native American genetic ancestry was ≥10% and was also greater than the proportion of African ancestry [8].

Results

The study cohort consists of a total of 733 individuals from 332 families: 287 families (628 individuals) from Unidad Nacional de Oncología Pediátrica (UNOP) in Guatemala, and 45 families (105 individuals) from Texas Children’s Hospital (TCH) in the United States. We excluded nine families from analysis because they did not meet Mendelian error rate criteria. The full cohort used in analyses included 323 families with 710 individuals. For 24 of these families, parental genetic data were used in the analysis, but genetic information was missing for the ALL cases themselves. Demographic and clinical characteristics of the 299 ALL cases included in this study are shown in Table 1. Of the 237,436 exonic SNPs available for analysis from the Illumina chip, 32,175 SNPs met the MAF and call rate criteria and were subsequently used in analyses (S1 File). The vast majority (99.6%) of excluded SNPs were omitted from analyses because they had a MAF of <1%.

[Figure omitted. See PDF.]

Table 1. Characteristics of ALL cases used in EXWAS (N = 299).

https://doi.org/10.1371/journal.pone.0180488.t001

In our exome-wide association analysis of the ALL cohort, we identified two SNPs on chromosome 10 within the ARID5B gene that were significantly associated with ALL risk: rs10821936 (RR = 2.31, 95% CI = 1.70–3.14, p = 1.70 × 10⁻⁸) and rs7089424 (RR = 2.22, 95% CI = 1.64–3.01, p = 5.19 × 10⁻⁸) (Fig 1 and Table 2). These two SNPs were in linkage disequilibrium (LD; r² = 0.84, D’ = 0.92) in the 1000 Genomes version hg19 Amerindian (AMR) reference population [31] (Fig 2). A subcohort analysis of only B-ALL cases and families (312 families, 683 individuals) yielded very similar effect estimates for rs10821936 (RR = 2.22, 95% CI = 1.63–3.02) and rs7089424 (RR = 2.13, 95% CI = 1.57–2.88), which were also significantly associated with B-ALL susceptibility (p = 9.63 × 10⁻⁸ and p = 2.81 × 10⁻⁷, respectively) (Table 2). Results for all SNPs that were nominally significant in our analysis (p<0.05) among the entire study cohort can be found in S2 Table.

[Figure omitted. See PDF.]

Table 2. Inherited effects results for ARID5B SNPs rs10821936 and rs7089424 (10q21.2), both for the entire ALL study cohort and for B-ALL cases and families alone.

https://doi.org/10.1371/journal.pone.0180488.t002

[Figure omitted. See PDF.]

Fig 1. Manhattan plot of inherited genetic effects in the full study cohort.

https://doi.org/10.1371/journal.pone.0180488.g001

[Figure omitted. See PDF.]

Fig 2. Regional association plot for ARID5B SNPs rs10821936 and rs7089424.

The left y-axis shows the multinomial model’s association probabilities (-log10p) for the SNP of interest (rs10821936; purple diamond) as well as for nearby SNPs used in the analysis. Red shading indicates linkage disequilibrium (r²) between the SNP of interest and nearby markers. Recombination rates for each region, obtained from 1000 Genomes build hg19 Amerindian (AMR) data, are indicated by blue lines measured on the right y-axis. The x-axis shows the position of the SNP of interest and other markers on the chromosome, relative to the position of nearby genes.

https://doi.org/10.1371/journal.pone.0180488.g002

Because this study included Hispanic childhood ALL cases and parents from two distinct geographic locations (Guatemala and Texas), analyses were, in effect, conducted on a pooled study population. We explored results for the ARID5B SNPs found to be of genome-wide significance in the overall study population in both of these populations separately, and effect sizes were consistent (rs10821936: RR = 2.32 for Guatemala, RR = 2.50 for Texas; rs7089424: RR = 2.27 for Guatemala, RR = 2.35 for Texas).

Discussion

In this family-based EXWAS of childhood ALL among those of Hispanic ancestry, we observed two inherited SNPs in ARID5B, rs10821936 and rs7089424, that were associated with childhood ALL. These two SNPs are in strong linkage disequilibrium with one another (r² = 0.84, D’ = 0.92), most likely representing a single susceptibility locus (Fig 2). Both ARID5B rs7089424 and rs10821936 have been identified in previous ALL GWAS [3–5,7,8]. These two SNPs also have had some of the strongest and most statistically significant associations with ALL in GWAS analyses [4,7,8]. We reviewed whole-exome sequence data from a subset of the cases (n = 41 from TCH) described here and did not identify any recurring ARID5B coding variants in patients with the risk allele.

We also compared findings from our analysis to susceptibility SNPs identified in previous GWAS/EXWAS of childhood ALL (Table 3). Notably, the magnitude of the effects of rs7089424 and rs10821936 in our study were over 20% higher than estimates calculated in previous genetic association studies among primarily European (non-Hispanic white) populations.

ARID5B SNPs have also been evaluated among Hispanics in particular in a small subset of studies. Our results are consistent with one other study, which observed a similar effect size for rs7089424 among Hispanics with childhood B-ALL (OR = 2.33, 95%CI: 1.85–2.92) [13].

[Figure omitted. See PDF.]

Table 3. Inherited effects reported in other ALL GWAS and EXWAS analyses, compared to our study results.

https://doi.org/10.1371/journal.pone.0180488.t003

It has been observed that these ARID5B variants appear to be more common among Hispanics [8,13]. As would be expected, the risk allele frequencies (RAFs) observed for both rs10821936 and rs7089424 among both cases and parents in our study were higher than the risk allele frequencies reported among Hispanic and Native American reference populations (S1 Table). However, a GWAS by Xu et al. (2013) observed a RAF of 0.63 for rs10821936 among Hispanic ALL cases [8], which was closer to the RAF observed among Hispanic ALL cases for this SNP in our study (0.73). It is possible that the increased frequency of the risk allele for both rs10821936 and rs7089424 among Hispanics and Native Americans could be a factor in the increased risk of childhood ALL among these populations.

The higher magnitude of effect observed among Hispanics in this study (as compared to studies among predominantly non-Hispanic whites) could be due to a stronger effect within this population for biological reasons, which are not apparent through this study. It is also possible that the larger effect size could be due to the study design used. The case-parent trio design may be a more unbiased approach in evaluating these effects, because it produces estimates that are inherently adjusted for effects of population stratification.

ARID5B regulates the transcription of certain genes during embryonic development, and plays a role in the differentiation of B-lymphocyte progenitor cells [7,32]. Therefore, germline variation at this locus could alter B-lymphocyte development, thus playing a part in susceptibility to B-ALL [7,33]. Furthermore, both rs7089424 and rs10821936 may have a role in transcriptional regulation. The risk allele for rs10821936, in particular, is thought to eliminate the binding site of transcriptional factor NIT2, which could alter gene expression in cis [34,35].

While SNPs in ARID5B were the only variants to reach genome-wide level significance in our study, previously-reported SNPs IKZF1 rs4132601 and rs11978267 were nominally significant (P<0.05), and their effect sizes were similar to previous assessments [3–5,7,8]. The only other SNPs previously found to be associated with ALL that were directly genotyped on the platform included PIP4K2A rs10828317 and rs7088318. Although PIP4K2A rs10828317 was not statistically significant, the effect size was comparable to that reported in Migliorini et al. (RR = 1.16 vs. 1.23) [11]. However, this was not the case for PIP4K2A rs7088318 (RR = 1.04 vs. 1.40) [8]. It is possible that variations in risk allele frequency among Hispanics could contribute to these differences, in part because prior assessments have largely focused on those of European ancestry. Because of the nature of the variants contained on the SNP array used in this study (i.e., the majority were less common exonic SNPs, with 86.1% of the SNPs on the chip having a MAF ≤1%), most of the SNPs associated with ALL in the literature were not available and therefore were not evaluated in this analysis.

This study should be considered in the light of certain limitations. As noted, because of the rare exonic content of the SNP chip, as well as the admixed study population, we did not perform SNP imputation. Additionally, because of the sample size, imputation of additional rare variants (and subsequent association testing) would have provided limited information. More specifically, our sample size did not allow for the detection of novel variants that are moderately associated with ALL. However, an important strength of our study was that it focused on Hispanic childhood ALL cases, in order to increase our knowledge of genetic risk factors among this higher-risk group. Populations of non-European ancestry are severely under-represented in GWAS; indeed, more than 95% of subjects in previous GWAS have been of European descent [19,36]. Because Hispanic children carry the burden of both a higher incidence of childhood ALL and worse outcomes of the disease than non-Hispanic whites or blacks [37,38], it is especially important to conduct exome- and genome-wide assessments for this disease among this particular ethnic group. Furthermore, a case-parent trio study design was used; unlike the case-control design employed by many GWAS, this approach is not subject to population stratification bias and is thus more likely to yield valid conclusions when assessing inherited genetic effects among Hispanic or other genetically diverse populations [16,17].

In summary, this is the first family-based study to evaluate inherited genetic variations associated with childhood ALL among those with Hispanic ancestry. These results confirm that ARID5B plays an important role in childhood ALL susceptibility in this high-risk population, and that the risk of childhood ALL associated with ARID5B rs10821936 and rs7089424 variants may be greater for Hispanics than among European populations. The study’s focus on Hispanic populations addresses an important gap in information about childhood ALL genetic risk factors for Hispanics.

Supporting information

[Figure omitted. See PDF.]

S1 Table. Risk allele frequencies (RAFs) of ARID5B rs10821936 and rs7089424 for each of the groups used in the analysis, as well as for selected reference populations.

https://doi.org/10.1371/journal.pone.0180488.s001

(DOCX)

S2 Table. Results for the 100 most highly significant SNPs in the inherited effects analysis, entire ALL study cohort.

https://doi.org/10.1371/journal.pone.0180488.s002

(DOCX)

S1 Fig. QC criteria based upon a combination of MAF and call rate (for those SNPs with a MAF <5%).

https://doi.org/10.1371/journal.pone.0180488.s003

(TIF)

S1 File. Compressed file folder containing allele frequencies for genotyped subjects and raw statistical results for SNPs included in analysis.

https://doi.org/10.1371/journal.pone.0180488.s004

(ZIP)

Acknowledgments

We would like to thank Larry Archer, Austin Brown, and Spiridon Tsavachidis for their technical assistance, as well as the families who participated in this study.

Citation: Archer NP, Perez-Andreu V, Stoltze U, Scheurer ME, Wilkinson AV, Lin T-N, et al. (2017) Family-based exome-wide association study of childhood acute lymphoblastic leukemia among Hispanics confirms role of ARID5B in susceptibility. PLoS ONE 12(8): e0180488. https://doi.org/10.1371/journal.pone.0180488

References

1. Pui C-H, Relling MV, Downing JR. Acute lymphoblastic leukemia. N Engl J Med. 2004;350:1535–48. pmid:15071128

2. Mullighan CG. The molecular genetic makeup of acute lymphoblastic leukemia. Hematology Am Soc Hematol Educ Program. 2012;2012:389–96. pmid:23233609

3. Ellinghaus E, Stanulla M, Richter G, Ellinghaus D, te Kronnie G, Cario G, et al. Identification of germline susceptibility loci in ETV6-RUNX1-rearranged childhood acute lymphoblastic leukemia. Leukemia. 2012;26:902–9. pmid:22076464

4. Orsi L, Rudant J, Bonaventure A, Goujon-Bellec S, Corda E, Evans TJ, et al. Genetic polymorphisms and childhood acute lymphoblastic leukemia: GWAS of the ESCALE study (SFCE). Leukemia. 2012 Dec;26:2561–4. pmid:22660188

5. Papaemmanuil E, Hosking FJ, Vijayakrishnan J, Price A, Olver B, Sheridan E, et al. Loci on 7p12.2, 10q21.2 and 14q11.2 are associated with risk of childhood acute lymphoblastic leukemia. Nat Genet. 2009 Sep;41:1006–10. pmid:19684604

6. Sherborne AL, Hosking FJ, Prasad RB, Kumar R, Koehler R, Vijayakrishnan J, et al. Variation in CDKN2A at 9p21.3 influences childhood acute lymphoblastic leukemia risk. Nat Genet. 2010;42:492–4. pmid:20453839

7. Trevino LR, Yang W, French D, Hunger SP, Carroll WL, Devidas M, et al. Germline genomic variants associated with childhood acute lymphoblastic leukemia. Nat Genet. 2009 Sep;41:1001–5. pmid:19684603

8. Xu H, Yang W, Perez-Andreu V, Devidas M, Fan Y, Cheng C, et al. Novel susceptibility variants at 10p12.31–12.2 for childhood acute lymphoblastic leukemia in ethnically diverse populations. J Natl Cancer Inst. 2013 May 15;105:733–42. pmid:23512250

9. Perez-Andreu V, Roberts KG, Harvey RC, Yang W, Cheng C, Pei D, et al. Inherited GATA3 variants are associated with Ph-like childhood acute lymphoblastic leukemia and risk of relapse. Nat Genet. 2013 Dec;45:1494–8. pmid:24141364

10. Perez-Andreu V, Roberts KG, Xu H, Smith C, Zhang H, Yang W, et al. A genome-wide association study of susceptibility to acute lymphoblastic leukemia in adolescents and young adults. Blood. 2015;125:680–6. pmid:25468567

11. Migliorini G, Fiege B, Hosking FJ, Ma Y, Kumar R, Sherborne AL, et al. Variation at 10p12.2 and 10p14 influences risk of childhood B-cell acute lymphoblastic leukemia and phenotype. Blood. 2013;122:3298–307. pmid:23996088

12. Yamamoto JF, Goodman MT. Patterns of leukemia incidence in the United States by subtype and demographic characteristics, 1997–2002. Cancer Causes Control. 2008;19:379–90. pmid:18064533

13. Walsh KM, Chokkalingam AP, Hsu LI, Metayer C, de Smith AJ, Jacobs DI, et al. Associations between genome-wide Native American ancestry, known risk alleles and B-cell ALL risk in Hispanic children. Leukemia. 2013 Dec;27:2416–9. pmid:23615557

14. Kadan-Lottick NS, Ness KK, Bhatia S, Gurney JG. Survival variability by race and ethnicity in childhood acute lymphoblastic leukemia. JAMA. 2003;290:2008–14. pmid:14559954

15. Yang JJ, Cheng C, Devidas M, Cao X, Fan Y, Campana D, et al. Ancestry and pharmacogenomics of relapse in acute lymphoblastic leukemia. Nat Genet. 2011 Mar;43:237–41. pmid:21297632

16. Cardon LR, Palmer LJ. Population stratification and spurious allelic association. Lancet. 2003;361:598–604. pmid:12598158

17. Zhao H. Family-based association studies. Stat Methods Med Res. 2000;9:563–87. pmid:11308071

18. Thornton TA, Bermejo JL. Local and global ancestry inference and applications to genetic association analysis for admixed populations. Genet Epidemiol. 2014 Sep;38 Suppl 1:S5–12.

19. Rosenberg NA, Huang L, Jewett EM, Szpiech ZA, Jankovic I, Boehnke M. Genome-wide association studies in diverse populations. Nat Rev Genet. 2010 May;11:356–66. pmid:20395969

20. Wilcox AJ, Weinberg CR, Lie RT. Distinguishing the effects of maternal and offspring genes through studies of “case-parent triads”. Am J Epidemiol. 1998;148:893–901. pmid:9801020

21. Archer NP, Perez-Andreu V, Scheurer ME, Rabin KR, Peckham-Gregory EC, Plon SE, et al. Family-based exome-wide assessment of maternal genetic effects on susceptibility to childhood B-cell acute lymphoblastic leukemia in Hispanics. Cancer. 2016;122:3697–704. pmid:27529658

22. Howey R, Cordell HJ. PREMIM and EMIM: tools for estimation of maternal, imprinting and interaction effects using multinomial modeling. BMC Bioinformatics. 2012;13:149. pmid:22738121

23. Weinberg CR, Wilcox AJ, Lie RT. A log-linear approach to case-parent-triad data: assessing effects of disease genes that act either directly or through maternal effects and that may be subject to parental imprinting. Am J Hum Genet. 1998;62:969–78. pmid:9529360

24. Ainsworth HF, Unwin J, Jamison DL, Cordell HJ. Investigation of maternal effects, maternal-fetal interactions and parent-of-origin effects (imprinting), using mothers and their offspring. Genet Epidemiol. 2011 Jan;35:19–45. pmid:21181895

25. Agopian AJ, Mitchell LE, Glessner J, Bhalla AD, Sewda A, Hakonarson H, et al. Genome-wide association study of maternal and inherited loci for conotruncal heart defects. PLoS One. 2014;9:e96057. pmid:24800985

26. Shi M, Murray JC, Marazita ML, Munger RG, Ruczinski I, Hetmanski JB, et al. Genome wide study of maternal and parent-of-origin effects on the etiology of orofacial clefts. Am J Med Genet A. 2012 Apr;158A:784–94. pmid:22419666

27. Mitchell LE, Agopian AJ, Bhalla A, Glessner JT, Kim CE, Swartz MD, et al. Genome-wide association study of maternal and inherited effects on left-sided cardiac malformations. Hum Mol Genet. 2015 Jan 1;24:265–73. pmid:25138779

28. Pruim RJ, Welch RP, Sanna S, Teslovich TM, Chines PS, Gliedt TP, et al. LocusZoom: regional visualiation of genome-wide association scan results. Bioinformatics. 2010;26:2336–7. pmid:20634204

29. Pritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genetics. 2000;155:945–59. pmid:10835412

30. Mao X, Bigham AW, Mei R, Gutierrez G, Weiss KM, Brutsaert TD, et al. A genomewide admixture mapping panel for Hispanic/Latino populations. Am J Hum Genet. 2007;80:1171–8. pmid:17503334

31. Ward LD, Kellis M. HaploReg: a resource for exploring chromatin states, conservation, and regulatory motif alterations within sets of genetically linked variants. Nucleic Acids Res. 2012;40:D930–934. pmid:22064851

32. Healy J, Richer C, Bourgey M, Kritikou EA, Sinnett D. Replication analysis confirms the association of ARID5B with childhood B-cell acute lymphoblastic leukemia. Haematologica. 2010 Sep;95:1608–11. pmid:20460642

33. Kennedy AE, Kamdar KY, Lupo PJ, Okcu MF, Scheurer ME, Dorak MT. Genetic markers in a multi-ethnic sample for childhood acute lymphoblastic leukemia risk. Leuk Lymphoma. 2015;56:169–74. pmid:24707947

34. Gutierrez-Camino A, Lopez-Lopez E, Martin-Guerrero I, Sanchez-Toledo J, Garcia de Andoin N, Carbone Baneres A, et al. Intron 3 of the ARID5B gene: a hot spot for acute lymphoblastic leukemia susceptibility. J Cancer Res Clin Oncol. 2013 Nov;139:1879–86. pmid:24013273

35. Lee PH, Shatkay H. F-SNP: computationally predicted functional SNPs for disease association studies. Nucleic Acids Res. 2008 Jan;36:D820–4. pmid:17986460

36. Bustamante CD, Burchard EG, De la Vega FM. Genomics for the world. Nature. 2011 Jul 14;475:163–5. pmid:21753830

37. Dores GM, Devesa SS, Curtis RE, Linet MS, Morton LM. Acute leukemia incidence and patient survival among children and adults in the United States, 2001–2007. Blood. 2012;119:34–43. pmid:22086414

38. Linabery AM, Ross JA. Trends in childhood cancer incidence in the U.S. (1992–2004). Cancer. 2008 Jan 15;112:416–32. pmid:18074355