Volume 14 Supplement 1
Critical Assessment of Data Privacy and Protection (CADPP)
Scalable privacy-preserving data sharing methodology for genome-wide association studies: an application to iDASH healthcare privacy protection challenge
- Fei Yu^{1}Email author and
- Zhanglong Ji^{2}
https://doi.org/10.1186/1472-6947-14-S1-S3
© Yu and Ji; licensee BioMed Central Ltd. 2014
Published: 8 December 2014
Abstract
In response to the growing interest in genome-wide association study (GWAS) data privacy, the Integrating Data for Analysis, Anonymization and SHaring (iDASH) center organized the iDASH Healthcare Privacy Protection Challenge, with the aim of investigating the effectiveness of applying privacy-preserving methodologies to human genetic data. This paper is based on a submission to the iDASH Healthcare Privacy Protection Challenge. We apply privacy-preserving methods that are adapted from Uhler et al. 2013 and Yu et al. 2014 to the challenge's data and analyze the data utility after the data are perturbed by the privacy-preserving methods. Major contributions of this paper include new interpretation of the χ^{2} statistic in a GWAS setting and new results about the Hamming distance score, a key component for one of the privacy-preserving methods.
Keywords
χ^{2} statistic Contingency table Differential privacy Genome-wide association study GWAS Data sharing Single-nucleotide polymorphismIntroduction
Rapid developments in whole-genome sequencing technologies in recent years have made the collection of high quality genetic data faster and more economically feasible. Many types of genetic research can benefit from having a large amount of genetic data. For example, in genome-wide association studies (GWAS), which are a type of genetic research that examine a large number of single-nucleotide polymorphisms (SNPs) to identify genetic factors associated with a phenotype, which is typically a common disease, increasing the number of DNA samples available for analysis allows researchers to make more accurate statistical inference and improve the overall quality of the analysis.
Encouraging data sharing among researchers is the first step towards taking advantage of the benefits brought about by the rapid growth in genetic data collection. However, being able to share genetic data without compromising the study participants' privacy remains one of the biggest challenges in genetic research. While it is clear that individual level genetic data deserve a high level of protection, for many years it was widely considered safe to release to the public aggregate genetic data pooled from thousands of individuals without compromising genetic study participants' privacy. However, Homer et al. [1] in 2008 demonstrated that one can use publicly available aggregate genetic data, such as SNP data from the International HapMap Project http://hapmap.ncbi.nlm.nih.gov/, to infer whether an individual has participated in a study. Cautious about the potential breach of genetic study participants' privacy, the National Institute of Health (NIH) quickly responded to the Homer et al. [1] attack by mandating an elaborate approval process that every researcher has to go through in order to gain access to aggregate genetic data. This NIH policy remains in effect today.
Homer et al. [1]'s attack and NIH's subsequent reaction spurred research interest in privacy-preserving methodologies for GWAS data. A recent concept of differential privacy (e.g. [2]), introduced by the cryptographic community, has shown great promise as a basis for privacy-preserving methodologies, as it provides a rigorous definition of privacy with meaningful privacy guarantees in the presence of arbitrary external information. We have seen privacy-preserving methods based on differential privacy applied to real human GWAS data in recent studies (e.g., [3–5]).
The iDASH Healthcare Privacy Protection Challenge, organized by Integrating Data for Analysis, Anonymization and SHaring (iDASH), aims to investigate the effectiveness of applying privacy-preserving methodologies to human genetic data [6]. This paper is based on a submission to the iDASH Healthcare Privacy Protection Challenge using privacy-preserving methods adapted from [3] and [5].
A major contribution of this paper is a new interpretation of the χ^{2} statistic in a GWAS setting and new results about the Hamming distance score, which plays an important role in the differentially private mechanisms proposed by [4] and [5]. In particular, we present a graphical interpretation of the allelic test χ^{2} statistic that will help us conceptualize the Hamming distance score. We also device an efficient algorithm for finding the Hamming distance score and prove that the sensitivity of the score function is 1; we hence address concerns raised in [5] about speed and sensitivity of alternative methods for finding the Hamming distance score.
We start by introducing background information on the iDASH Healthcare Privacy Protection Challenge. We briefly describe the characteristics of the data and define the allelic test χ^{2} statistic, which is used for evaluating the performance of submissions in the challenge. Then we summarize differentially private mechanisms applied to the challenge's data, which include a mechanism based on the Laplace mechanism and χ^{2} statistic, a mechanism based on the exponential mechanism and χ^{2} statistic, and a mechanism based on the exponential mechanism and Hamming distance score. We present a graphical interpretation of the allelic test χ^{2} statistic and an efficient algorithm for finding the Hamming distance score. We prove that our algorithm finds the shortest Hamming distance and therefore the Hamming distance score has sensitivity 1. We incorporate our improvements into the differentially private mechanisms and apply them to the challenge's data. We compare the performance of the mechanisms using risk-utility plots.
Background information on iDASH challenge
The challenge has two tasks, both of which are concerned with the dissemination of aggregate GWAS data: (1) limiting the re-identification risks when releasing all aggregate data in a GWAS dataset, and (2) being compliant with differential privacy (Definition 2) when releasing the most significant SNPs. This paper focuses on the second task of releasing the most significant SNPs differentially privately.
The data used for the second task consist of 201 participants from the Personal Genome Project (http://www.personalgenomes.org/) and 174 participants from HapMap. Individuals from PGP are treated as cases and those from HapMap are treated as controls in the challenge. 106,129 SNPs are typed in all participants. [6] has more details on how the data are processed.
A subset containing 5,000 SNPs is selected by organizers of the challenge to form a representative sample of the entire set of SNPs. This paper uses the subset of SNPs to evaluate the performance of the privacy-preserving methods, as is recommended by organizers of the challenge.
Genotype table
# of minor alleles | Total | |||
---|---|---|---|---|
0 | 1 | 2 | ||
Case | r _{0} | r _{1} | r _{2} | R |
Control | s _{0} | s _{1} | s _{2} | S |
Total | n _{0} | n _{1} | n _{2} | N |
Allelic table
Allele type | Total | ||
---|---|---|---|
Minor | Major | ||
Case | r_{1} + 2r_{2} | 2r_{0} + r_{1} | 2R |
Control | s_{1} + 2s_{2} | 2s_{0} + s_{1} | 2S |
Total | n_{1} + 2n_{2} | 2n_{0} + n_{1} | 2N |
In this challenge, the statistical significance of a SNP's association with the phenotype is assessed by the the allelic test statistic (Definition 1). For the rest of the paper, we will simply refer to the allelic test statistic as χ^{2} statistic. Assuming that the control group's data are public, we will use the differentially private mechanisms discussed in the next section to release the top K SNPs while preserving the privacy of the case group.
Differential privacy: definitions and methods
The concept of differential privacy, recently introduced by the cryptographic community (e.g., [2]), provides a notion of privacy guarantees that protect GWAS databases against arbitrary external information.
Two methods are often used as building blocks for constructing more complex differentially private algorithms. One of the methods, due to [2], is called the Laplace mechanism (Definition 4), and the other method, due to [7], is called the exponential mechanism (Definition 5). Both methods require knowledge of the sensitivity of the score function, where sensitivity is defined as the smallest upper bound of how much the function can vary when one record in the input database changes (see Definition 3).
for all databases $D,{D}^{\prime}\in \mathcal{D}$such that D ~ D'.
Definition 4 (Laplace mechanism) Releasing f (D)+b, where b ~ Laplace $\left(0,\frac{S\left(f\right)}{\epsilon}\right)$, satisfies the definition of ∈-differential privacy.
Then releasing ${\epsilon}_{q}^{\epsilon}$ satisfies the definition of ∈-differential privacy.
Methods for releasing the Kmost relevant SNPs
Algorithm 1 The E-differentially private mechanism for releasing the K most relevant SNPs using the Laplace mechanism [3, 5, 8].
Input: The score of all M candidate SNPs, the number of SNPs, K, that we want to release, the sensitivity, s, of the score function, and the privacy budget E.
Output: K SNPs.
1: Add independent Laplace noise with mean zero and scale $\frac{2K\phantom{\rule{0.3em}{0ex}}s}{\epsilon}$ to each of the M SNPs scores.
2: Choose the top K SNPs based on the perturbed scores.
Algorithm 2 The ∈-differentially private mechanism for releasing the K most relevant SNPs using the exponential mechanism [4, 5].
Input: The scores (e.g. χ^{2} statistic or Hamming distance) of all M candidate SNPs, the number of SNPs, K, that we want to release, the sensitivity, s, of the score function, and the privacy budget ∈.
Output: K SNPs.
1: Initialize ${\left\{{q}_{i}\right\}}_{i=1}^{M}$ score of SNP_{ i }.
2: Set ${w}_{i}=\mathsf{\text{exp}}\left(\frac{\epsilon {q}_{i}}{2Ks}\right)$. Define $\mathsf{\text{Pr}}\left(\mathcal{T}\left(D\right)=i\right)={w}_{i}/{\displaystyle \sum _{j=1}^{M}}{w}_{j}.$
3: Sample $j~\mathcal{T}\left(D\right)$. Record SNP_{ j } . Set q_{ j } = −∞.
4: Repeat Step 2 and 3 until K SNPs have been recorded.
Algorithm 1 and Algorithm 2 extend the Laplace mechanism and the exponential mechanism, respectively, to release more than a single SNP differentially privately. In this paper, we consider three mechanisms for releasing the top K SNPs: a mechanism that is based on Algorithm 1 and uses χ^{2} statistic as score function, a mechanism that is based on Algorithm 2 and uses χ^{2} statistic as score function, and a mechanism that is based on Algorithm 2 and uses the Hamming distance score ([4]) as score function. In loose terms, the Hamming distance score is the smallest number of changes made to a genotype table until the significance of the table changes, where a change, counted as 1-Hamming distance in the space of genotype tables, is defined as changing the genotype of one individual and significance refers to whether the p-value of the χ^{2} statistic of the table is smaller than a pre-specified threshold value or not. See [5] for more details on the three mechanisms and applications of them to a real human GWAS dataset.
For mechanisms that use the χ^{2} statistic as score, we need to know the sensitivity of the χ^{2} statistic. An upper bound for the sensitivity is shown in [5], but [5] requires that the margins of the genotype contingency tables to be positive. Indeed, such requirement can be satisfied in the challenge's setting when we assume that Hardy-Weinberg equilibrium holds: because a typical GWAS dataset consists of only common SNPs, whose minor allele frequencies are greater than 1%, the control group's three genotypes derived from the allele frequency at each SNPs will be nonnegative, which ensures that the derived genotype tables have positive margins.
For the mechanism that uses the Hamming distance score as score, we already know that, by construction, the sensitivity of the score function is 1 if the Hamming distance is the shortest Hamming distance [4]. However, as is pointed out in [4] and [5], it is a computationally onerous task to actually calculate the shortest Hamming distance, which, in the most naïve setting, involves examining all possible sequential changes made to the original genotype table that alter the significance status of the table. To make the calculations more computationally feasible, [4] and [5] use approximations of the shortest Hamming distance in their implementations of the mechanism, noting the caveat that the sensitivity of the approximated Hamming distance score may no longer be 1.
In the next section, we propose a new method of finding the Hamming distance score that is much more computationally efficient than those in [4] and [5]. We also prove that our method indeed produces the shortest Hamming distance, and therefore the sensitivity of the resulting Hamming distance score function is 1.
Finding the Hamming distance score
Let p^{*} denote a pre-specified threshold p-value and let c denote the χ^{2} statistic corresponding to p^{ ∗ }, the p-value of the χ^{2} distribution with 1 degree of freedom. Then for a given SNP in the pool of candidate SNPs, the genotype table of which we denote by D, the shortest Hamming distance is the smallest number of sequential changes made to D such that the resulting genotype table, D', satisfies Y_{ A }(D') ≥ c if Y_{ A }(D) < c and Y_{ A }(D') < c if Y_{ A }(D) ≥ c; that is, if we call c the significance threshold, then the goal is to make changes to the "insignificant" ("significant") table D so that the χ^{2} statistic of D' goes above (below) the significance threshold c, and D' becomes a "significant" ("insignificant") table. The Hamming distance score is defined as h = (shortest Hamming distance) − 1 if Y_{ A }(D) ≥ c and h = −(shortest Hamming distance) if Y_{ A }(D) < c.
where r_{0}, r_{1} and x are derived from D', and n_{10}, R, S and N are the same for D and D^{ l }. For notational convenience, when r_{0} and r_{1} are also derived from D, we will simply write the χ^{2} statistic as Y_{ A }(D).
Lemma 1 Y_{ A } is an increasing function of × when xS − n_{10}R >0, and it is a decreasing function of × when xS − n_{10}R <0.
Proof. [see Additional file 1].
there could be genotype tables for which only one black line exists or no black line exists at all; in such cases, we will use the lines 0 = 2r_{0} + r_{1} or 2R = 2r_{0} + r_{1} wherever appropriate.
In Figure 2, the genotype table D is insignificant and its χ^{2} statistic is below the threhold value. By Lemma 1, we know that the χ^{2} statistics of genotype tables, as represented by the dots on Figure 2, are greater than c when they are in the shaded area, outside of the area between the two black lines and they are smaller than Y_{ A }(D^{ ∗ }) when they are inside the area between the two black lines. Therefore, finding the Hamming distance score for D is to find the shortest Hamming distance from the genotype table D to genotype tables in the shaded areas.
For genotype tables that are significant, they will fall into the shaded areas in Figure 2. Then finding the Hamming distance score for a significant genotype table is to find the shortest Hamming distance from the genotype table in one of the shaded areas to genotype tables in the non-shaded area.
Proposition 2 Given a significance threshold value c and an insignificant genotype table D (i.e., Y_{ A }(D) < c), if there exists ${D}^{\prime}\in {\mathcal{B}}_{D}$ such that ${Y}_{A}\left({D}^{\prime};{\mathcal{B}}_{D}\right)\ge c$, then the shortest Hamming distance is min{H_{1}, H_{2}}, where H_{1} and H_{2} are defined as follows:
(i) H_{1} is the number of changes made to D in the following manner: (1) keep decreasing r_{0} until the new genotype table, D', becomes significant (i.e.,${Y}_{A}\left({D}^{\prime};\mathcal{D}\right)>c$); (2) when r_{0} is minimized but the new table is still insignificant, keep decreasing r_{1} until the new table becomes significant.
(ii) H_{2} is the number of changes made to D in the following manner: (1) keep increasing r_{0} until the new genotype table becomes significant; (2) if r_{0} can no longer be increased without decreasing r_{1} and the new table is still insignificant, increase r_{0} and decrease r_{1} in each change until the new table becomes significant.
If for all ${D}^{\prime}\in {\mathcal{B}}_{D}$, ${Y}_{A}\left({D}^{\prime};{\mathcal{B}}_{D}\right)<c$, then we define the shortest Hamming distance as min$\left\{{H}_{1}^{\prime},{H}_{2}^{\prime}\right\}$, where ${H}_{1}^{\prime}$ and ${H}_{2}^{\prime}$ are defined as follows:
(i) When r_{0} and r_{1} are both minimized but the new table is still insignificant, set ${H}_{1}^{\prime}$ to 1 + d_{1}, where d_{1} is smallest the number of changes needed to minimize r_{0} and r_{1}.
(ii) When r_{0} and r_{1} are both maximized but the new table is still insignificant, set ${H}_{2}^{\prime}$ to 1 + d_{2}, where d_{2} is smallest the number of changes needed to maximize r_{0} and r_{1}.
Proof. [see Additional file 1].
Proposition 3 Given a significance threshold value c and a significant genotype table D (that is, Y_{ A }(D) ≥ c), the shortest Hamming distance is min{H_{1}, H_{2}}, where H_{1} and H_{2} are defined as follows:
(i) If 2r_{0} + r_{1} >(2s_{0} + s_{1})R/S, set H_{1} = ∞; otherwise, H_{1} is the number of changes made to D in the following manner: keep decreasing r_{0} until the new genotype table, D', becomes insignificant (i.e., Y_{ A }(D', D) < c).
(ii) If 2r_{0} + r_{1} <(2s_{0} + s_{1})R/S, set H_{2} = ∞; otherwise, H_{2} is the number of changes made to D in the following manner: keep decreasing r_{0} until the new genotype table becomes insignificant.
Proof The proof is similar to that of Proposition 2.
where d^{ − } is found using Proposition 2 and d^{+} is found using Proposition 3.
Corollary 4 The sensitivity of the Hamming distance score as defined in Definition 6 is 1.
Application to the challenge's data
It is also worth noting that even though the performance of the mechanism based on the Hamming distance score does not seem to be affected by the choice of threshold p-value, the analysis of the mechanism in [5] shows that whether the choice of threshold p-value has any effect on data utility also depends on the choice of K, the number of top SNPs to release. Therefore, the choice of threshold p-value should be justified before we use this mechanism.
Conclusions
In our submission to the iDASH Healthcare Privacy Protection Challenge, we apply differentially-private methods proposed by [3] and [5] to the challenge's data. Our results show that the performance of the method based on Algorithm 2 and Hamming distance score is superior to that of other methods when the privacy budget, ∈, is small. But we also point out problems with the Hamming distance score, such as the data utility plateauing at a level lower than other methods.
We devise an efficient algorithms for finding the Hamming distance score and prove that the sensitivity of the score function is 1. This addresses concerns raised in [5] regarding speed and sensitivity of alternative methods for finding the Hamming distance score. The graphical interpretation of the χ^{2} statistic that we present in the paper is instrumental in our discovery of the efficient algorithm for finding the Hamming distance score. We expect that the graphical interpretation can be extrapolated to other settings, such as the Pearson's χ^{2} statistic for 2 × 3 contingency tables and the setting in which data for the controls are not assumed to be public, and help with designing efficient algorithms for fining the Hamming distance score in those settings.
Declarations
Acknowledgements
This research was partially supported by NSF Awards EMSW21-RTG and BCS-0941518 to the Department of Statistics at Carnegie Mellon University.
Publication of this article has been funded by iDASH(U54HL108460), NHGRI (K99HG008175), NLM(R00LM011392,R21LM012060), CTSA(UL1TR000100), and an NCBC-linked grant (R01HG007078).
This article has been published as part of BMC Medical Informatics and Decision Making Volume 14 Supplement 1, 2014: Critical Assessment of Data Privacy and Protection (CADPP). The full contents of the supplement are available online at http://www.biomedcentral.com/bmcmedinformdecismak/supplements/14/S1.
Authors’ Affiliations
References
- Nils Homer, Szelinger Szabolcs, Redman Margot, Duggan David, Tembe Waibhav, Muehling Jill, Pearson John, Stephan Dietrich, Nelson Stanley, Craig David: Resolving individuals contributing trace amounts of DNA to highly complex mixtures using high-density SNP genotyping microarrays. PLoS Genetics. 2008, 4 (8): e1000167-10.1371/journal.pgen.1000167.View ArticleGoogle Scholar
- Dwork Cynthia, McSherry Frank, Nissim Kobbi, Smith Adam: Calibrating noise to sensitivity in private data analysis. Theory of Cryptography. 2006, 1-20.Google Scholar
- Uhler Caroline, Slavkovic Aleksandra, Fienberg Stephen: Privacy-preserving data sharing for genome-wide association studies. Journal of Privacy and Confidentiality. 2013, 5 (1): 137-166.Google Scholar
- Johnson Aaron, Shmatikov Vitaly: Privacy-preserving data exploration in genome-wide association studies. Proceedings of the 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 2013, 1079-1087.View ArticleGoogle Scholar
- Yu Fei, Fienberg Stephen, Slavković Aleksandra, Uhler Caroline: Scalable privacy-preserving data sharing methodology for genome-wide association studies. Journal of biomedical informatics. 2014, 50C: 133-141.View ArticleGoogle Scholar
- Jiang Xiaoqian, Zhao Yongan, Wang Xiaofeng, Malin Bradley, Wang Shuang, Ohno-Machado Lucila, Tang Haixu: A community assessment of privacy preserving techniques on human genome data. BMC. 2014Google Scholar
- McSherry Frank, Talwar Kunal: Mechanism Design via Differential Privacy. 48th Annual IEEE Symposium on Foundations of Computer Science (FOCS'07). 2007, 94-103.View ArticleGoogle Scholar
- Bhaskar Raghav, Laxman Srivatsan, Smith Adam, Thakurta Abhradeep: Discovering frequent patterns in sensitive data. Proceedings of the 16th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - KDD '10. 2010, New York, New York, USA, ACM Press, 503-View ArticleGoogle Scholar
Copyright
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/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.