 Short Report
 Open Access
 Published:
The effect of alternative permutation testing strategies on the performance of multifactor dimensionality reduction
BMC Research Notesvolume 1, Article number: 139 (2008)
Abstract
Background
Multifactor Dimensionality Reduction (MDR) is a novel method developed to detect genegene interactions in casecontrol association analysis by exhaustively searching multilocus combinations. While the endgoal of analysis is hypothesis generation, significance testing is employed to indicate statistical interest in a resulting model. Because the underlying distribution for the null hypothesis of no association is unknown, nonparametric permutation testing is used. Lately, there has been more emphasis on selecting all statistically significant models at the end of MDR analysis in order to avoid missing a true signal. This approach opens up questions about the permutation testing procedure. Traditionally omnibus permutation testing is used, where one permutation distribution is generated for all models. An alternative is nlocus permutation testing, where a separate distribution is created for each nlevel of interaction tested.
Findings
In this study, we show that the false positive rate for the MDR method is at or below a selected alpha level, and demonstrate the conservative nature of omnibus testing. We compare the power and false positive rates of both permutation approaches and find omnibus permutation testing optimal for preserving power while protecting against false positives.
Conclusion
Omnibus permutation testing should be used with the MDR method.
Background
One of the main goals of genetic epidemiology is the identification and characterization of polymorphisms that present an increased risk of disease. It is increasingly assumed that complex diseases are the result of a myriad of genetic and environmental risk factors [1, 2]. This complex etiology limits the utility of traditional, parametric statistical approaches in genetic association studies [3, 4]. The ubiquitous nature of genegene and geneenvironment interactions [1, 5, 6] has inspired the development the novel statistical approaches designed to detect epistasis [7–9].
Multifactor Dimensionality Reduction (MDR) is one such method [10]. MDR was designed to detect interactions in categorical independent variables and a dichotomous dependent variable (i.e. case/control status or drug treatment response/nonresponse). MDR performs an exhaustive search of all possible singlelocus through nlocus interactions (as computationally feasible) to evaluate all possible high/low risk models of disease. MDR selects a single model as optimal for each nlocus interaction as a result of these evaluations. Permutation testing (PT) is used to determine the significance of these models. MDR is nonparametric and modelfree, so no hypotheses concerning the value of any statistical parameter nor any genetic inheritance model are made [10]. MDR has successfully identified interactive effects in simulated data as well as real data applications in diseases such as hypertension [3, 11, 12], cancer [10, 13, 14], and atrial fibrillation [15, 16].
The endgoal of an MDR analysis is ultimately hypothesis generation (or refinement within candidate gene strategies) [17]. Hypothesis testing is used within the MDR analysis framework to determine whether resulting models are significantly different than expected by chance. Significance of a model is intended to indicate an interesting model that should be followed up in replication cohorts or functional studies. In recent work, there has been more emphasis on selecting all statistically significant models [17] in order to avoid missing a true signal (false negatives) in exchange for risking the selection of a few false positives. This generation of multiple hypotheses opens up questions about the PT procedure used to ascribe significance to this end set of models.
PT is a commonly used nonparametric statistical procedure that involves resampling the data without replacement to actually construct the distribution of the test statistic under the null hypothesis rather than make specific distributional assumptions. If the value of the test statistic based on the original samples is extreme relative to this distribution (i.e. if it falls far into the tail of the distribution), then the null hypothesis is rejected [18]. Validity of PT relies only on the property of exchangeability under the null hypothesis – that the joint distribution of the data samples must remain invariant to permutations of the data subscripts. Thus, permutation tests maintain a wide applicability under a much broader range of data and research conditions than most parametric tests [19]. In addition, PT requires minimal assumptions about the data being examined, yet often has power equal to, or even greater than, parametric counterparts that require stronger, and sometimes untenable data assumptions [20]. Unlike many parametric and other nonparametric tests, the results of permutation tests (the pvalues) are unbiased [18]. The chief drawback of this method is that it is computationally expensive, but the easy availability of fast computing has made this a practical approach even for large datasets.
MDR implements PT to statistically test to the best model(s) [21]. Typically, omnibus PT is used, where a single null distribution is generated from the best model of each of at least one thousand randomized datasets. With a focus on selecting all potentially interesting models from the final MDR set, this omnibus method may be too conservative. nlocus PT is an alternative, where a separate null distribution is created for each nlevel of interaction. So if singlelocus through fiveway interactions were evaluated in an original MDR analysis, a separate distribution would be created for the singlelocus model, for the twolocus model, etc (for a total of five null distributions).
Currently, we compare the significance cutoffs, power, and false positive rates of omnibus PT and nlocus PT implemented in MDR for a wide range of disease models. We also examine the overall false positive rate of the MDR method using both types of PT. As the MDR method gains acceptance and is increasingly used in the genetics community, it is important that users understand how to properly apply PT.
Multifactor Dimensionality Reduction (MDR)
Figure 1 (adapted from [10]) outlines the MDR procedure. Details of the algorithm and of the alternative PT strategies implemented in the current study can be found in Additional file 1.
Data Simulations and Analysis
Simulated datasets that exhibit genegene interactions were generated for the purpose of evaluating the power and false positives of MDR using either omnibus or nlocus PT. Multiple disease models, as well as null data with no disease model, were generated with varying allele frequencies, heritability, and number of interacting functional polymorphisms. Details of the simulations and analysis are found in Additional file 1.
Results
Null data was used to check the false positive rates of both permutationtesting strategies in the absence of any signal from the data. Best models were chosen for each dataset based on low prediction error and high crossvalidation consistency and compared to the appropriate permutation distribution. Table 1 shows the false positive rate for each permutation distribution, with alpha = 0.05, where the false positive rate is estimated as the number of models that were declared significant by PT out of the 100 datasets analyzed. These results demonstrate that the false positive rate is nominal for each permutation distribution – the error rate is at or below the selected alpha level.
Both omnibus and nlocus permutation distributions were created for each model, and the highest prediction error that would be ascribed statistical significance at the alpha = 0.05 level was recorded (cutoffs for significance). Table 2 lists these cutoffs for omnibus testing and each possible nlocus distribution. For each model, the most conservative cutoff is highlighted with bold font. These results demonstrate that omnibus PT consistently provides the most conservative PT cutoffs, as its cutoff prediction errors are the lowest. There is also a general trend within the nlevel PT distributions that as the level of n increases, so does the corresponding cutoff value. This demonstrates that the nlocus PT becomes more liberal as the level of dimensionality increases.
While the anticonservative nature of nlocus PT could potentially increase power, it is undesirable if that results in an increased false positive rate. To evaluate this, we investigated the false positive rate of each nlocus permutation distribution for each model. MDR analysis was performed on each dataset for all singlelocus through fivelocus combinations, and a best model was chosen for each level of interaction. For each dataset, the best model for each level of interaction was compared to the appropriate nlocus permutation distribution to estimate the false positive rate where a false positive result was any model that is not correct (may contain only incorrect loci or correct loci with additional false positive loci) and was found statistically significant according to the appropriate permutation distribution. Summarized in Table 3, using this definition of power, the false positive rate is extremely high for any nlevel interaction above the true genetic model. For example, for the twolocus interaction model with 0.2 minor allele frequency and 3% heritability, all three, four, and five locus models were statistically significant.
To better understand this trend, we estimated power for each model as the number of times all functional loci (with or without additional/false positive loci) were identified within the best model for any nlevel interaction and was called significant according to the corresponding permutation distribution out of the 100 simulated datasets per model. Table 4 summarizes these results. These results suggest that the false positive rates shown in Table 3 may be driven by the inclusion of functional loci in higherlevel interactions. This trend was seen for all models, but is especially apparent in higher heritability models. This suggests, especially in the case of a relatively strong signal from the data, that even containing the correct loci within the model is enough to drive it to statistical significance according to a more liberal PT procedure. This is a highly likely explanation, especially considering the nominal false positive rates demonstrated for null data (Table 1).
Table 4 shows that the power of MDR is relatively high in lower order models, especially at the nlocus level of analysis. Interpretation of an MDR analysis is complicated, however, when using nlocus PT by the high level of false positives. Even though the functional loci are included in the significant models, choosing the correct order of interaction is difficult when nlocus PT is used for each level of interaction.
Understanding that omnibus PT is the more conservative option, we investigated its impact on both the overall power and false positive rates of the MDR method. Table 5 summarizes these results for each disease model. First, we wanted to compare the power of MDR to detect the correct model as the final model (through minimization of prediction error and maximization of cross validation consistency) without considering statistical significance. Power was estimated as the number of times the functional/disease associate loci were chosen as the best model across the 100 replicates, with no false positive loci included in the model. By defining power in this context (with no PT), this estimate represents the least conservative estimate. These results are equivalent to the least conservative cutoff possible with nlocus PT. By defining power in this way, by using any significance testing the power cannot possibly be higher – all results that count towards "power" under this definition can only be changed to nonsignificant by using any significance testing. Results are summarized in the column of Table 5 labeled "Power Without Permutation Testing". This is then compared to the power of MDR to not only find the correct model, but to also ascribe statistical significance to that model through omnibus PT. Results of this evaluation are presented in the column of Table 5 labeled "Power With Permutation Testing". Comparing the power with and without permutation demonstrates the results are similar. Omnibus PT does not severely limit the power of the method.
Finally, we evaluated the impact of omnibus PT on the false positive rate. The false positive rate was estimated for each model as the number of incorrect final models that were statistically significant using omnibus PT. This calculation included significance testing at each level of interaction – not just a single test for one overall best model. As Table 5 shows, the false positive rates are near the expected 5% level.
From the results presented in Table 5, we conclude that omnibus PT controls for false positives while preserving power.
Conclusion
In this study we confirmed that the overall false positive rate of MDR is as expected according to the selected alpha level. Additionally, we demonstrated the conservative nature of omnibus testing in comparison to an nlocus strategy.
We also demonstrated that omnibus PT is preferred to nlocus since it controls false positives without limiting power. While MDR has high power using either permutationtesting scenario, final model selection is complicated by the more liberal nlocus strategy of PT. While the final goal of MDR is hypothesis generation, and the user may prefer the risk of false positives to the risk of missing a true signal, it is recommended that significance levels be assigned to one or more models from the final set using the omnibus permutation distribution, and not using corresponding nlocus tests.
While these results are most immediately applicable to genetic epidemiologists using MDR, they may generalize to any computational method that involves PT. Additionally, as MDR gains acceptance and becomes more widely used, it is important that the consequences of alternative permutation strategies should be explored and understood. Recent work is also implementing alternative hypothesis testing strategies for MDR that are computationally feasible for extremely largescale datasets [22].
Abbreviations
 MDR:

Multifactor Dimensionality Reduction
 PT:

Permutation testing.
References
 1.
Moore JH: The ubiquitous nature of epistasis in determining susceptibility to common human diseases. Hum Hered. 2003, 56: 7382.
 2.
Templeton A: Epistasis and complex traits. Wade, M., Broadie, B III, and Wolf, J. 41–57. Epistasis and the Evolutionary Process. 2000, Oxford, Oxford University Press
 3.
Moore JH, Williams SM: New strategies for identifying genegene interactions in hypertension. Ann Med. 2002, 34: 8895.
 4.
Moore JH: Computational analysis of genegene interactions using multifactor dimensionality reduction. Expert Rev Mol Diagn. 2004, 4: 795803.
 5.
Sing CF, Stengard JH, Kardia SL: Dynamic relationships between the genome and exposures to environments as causes of common human diseases. World Rev Nutr Diet. 2004, 93: 7791.
 6.
ThorntonWells TA, Moore JH, Haines JL: Genetics, statistics and human disease: analytical retooling for complexity. Trends Genet. 2004, 20: 640647.
 7.
Culverhouse R, Klein T, Shannon W: Detecting epistatic interactions contributing to quantitative traits. Genet Epidemiol. 2004, 27: 141152.
 8.
Millstein J, Conti DV, Gilliland FD, Gauderman WJ: A testing framework for identifying susceptibility genes in the presence of epistasis. Am J Hum Genet. 2006, 78: 1527.
 9.
Nelson MR, Kardia SL, Ferrell RE, Sing CF: A combinatorial partitioning method to identify multilocus genotypic partitions that predict quantitative trait variation. Genome Res. 2001, 11: 458470.
 10.
Ritchie MD, Hahn LW, Roodi N, Bailey LR, Dupont WD, Parl FF, et al: Multifactordimensionality reduction reveals highorder interactions among estrogenmetabolism genes in sporadic breast cancer. Am J Hum Genet. 2001, 69: 138147.
 11.
Sanada H, Yatabe J, Midorikawa S, Hashimoto S, Watanabe T, Moore JH, et al: Singlenucleotide polymorphisms for diagnosis of saltsensitive hypertension. Clin Chem. 2006, 52: 352360.
 12.
Williams SM, Ritchie MD, Phillips JA, Dawson E, Prince M, Dzhura E, et al: Multilocus analysis of hypertension: a hierarchical approach. Hum Hered. 2004, 57: 2838.
 13.
Andrew AS, Nelson HH, Kelsey KT, Moore JH, Meng AC, Casella DP, et al: Concordance of multiple analytical approaches demonstrates a complex relationship between DNA repair gene SNPs, smoking and bladder cancer susceptibility. Carcinogenesis. 2006, 27: 10301037.
 14.
Xu J, Lowey J, Wiklund F, Sun J, Lindmark F, Hsu FC, et al: The interaction of four genes in the inflammation pathway significantly predicts prostate cancer risk. Cancer Epidemiol Biomarkers Prev. 2005, 14: 25632568.
 15.
Motsinger AA, Donahue BS, Brown NJ, Roden DM, Ritchie MD: Risk factor interactions and genetic effects associated with postoperative atrial fibrillation. Pac Symp Biocomput. 2006, 584595.
 16.
Tsai CT, Lai LP, Lin JL, Chiang FT, Hwang JJ, Ritchie MD, et al: Reninangiotensin system gene polymorphisms and atrial fibrillation. Circulation. 2004, 109: 16401646.
 17.
Motsinger AA, Ritchie MD: Multifactor dimensionality reduction: an analysis strategy for modelling and detecting genegene interactions in human genetics and pharmacogenomics studies. Hum Genomics. 2006, 2: 318328.
 18.
Good P: Permutation Tests: A Practical Guide to Resampling Methods for Testing Hypotheses. 2000, New York, SpringerVerlag
 19.
Mukherjee S, Golland P, Panchenko D: Permutation tests for classification. AI Memo #2003019. 2003, Cambridge MA, Massachusetts Institure of Technology
 20.
Neuhauser M: Efficiency comparisons of rank and permutation tests. Stat Med. 2005, 24: 17771778.
 21.
Hahn LW, Ritchie MD, Moore JH: Multifactor dimensionality reduction software for detecting genegene and geneenvironment interactions. Bioinformatics. 2003, 19: 376382.
 22.
Pattin KA, White BC, Barney M, Gui J, Nelson HH, Kelsey KT: A computationally efficient hypothesis testing method for epistasis analysis using multifactor dimensionality reduction. Genet Epidemiol.
Acknowledgements
This work was partially supported by National Institutes of Health grant GM62758. We would like to thank Dr. Marylyn Ritchie for helpful discussion on the manuscript.
Author information
Additional information
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
AAM designed the study, performed the data analysis, and wrote the manuscript.
Authors’ original submitted files for images
Below are the links to the authors’ original submitted files for images.
Rights and permissions
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
About this article
Received
Accepted
Published
DOI
Keywords
 False Positive Rate
 Permutation Testing
 Multifactor Dimensionality Reduction
 Permutation Distribution
 Multifactor Dimensionality Reduction Analysis