Resolving Individuals Contributing Trace Amounts of DNA to Highly Complex Mixtures Using High-Density SNP Genotyping Microarrays

By Nils Homer, S. Szelinger, Margot Redman, D. Duggan, Waibhav Tembe, Jill Muehling, J. Pearson, D. Stephan, Stanley F. Nelson, D. Craig. 2008

1,347 citations recorded by the source corpus.

We use high-density single nucleotide polymorphism (SNP) genotyping microarrays to demonstrate the ability to accurately and robustly determine whether individuals are in a complex genomic DNA mixture. We first develop a theoretical framework for detecting an individual's presence within a mixture, then show, through simulations, the limits associated with our method, and finally demonstrate experimentally the identification of the presence of genomic DNA of specific individuals within a series of highly complex genomic mixtures, including mixtures where an individual contributes less than 0.1% of the total genomic DNA. These findings shift the perceived utility of SNPs for identifying individual trace contributors within a forensics mixture, and suggest future research efforts into assessing the viability of previously sub-optimal DNA sources due to sample contamination. These findings also suggest that composite statistics across cohorts, such as allele frequency or genotype counts, do not mask identity within genome-wide association studies. The implications of these findings are discussed.

Subfields: Genetics, Evaluation & Benchmarks

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Genetics

In Evaluation & Benchmarks

Citations in the corpus

Privacy-Preserving Generative Deep Neural Networks Support Clinical Data Sharing (2017, background): “showed the ability to identify whether an individual was a part of a study even with complex genomic mixtures (22).”
Membership Inference Attacks Against Machine Learning Models (2016, methodology): “By contrast, the membership inference problem we study in this paper is essentially the same as the well-known problem of identifying the presence of an individual’s data in a mixed pool given some statistics about the pool [3], [15], [21], [29].”
Model Inversion Attacks that Exploit Confidence Information and Basic Countermeasures (2015, background): “[18] consider disclosure attacks on datasets generated from Genome-Wide Association Studies (GWAS) and other partial genetic information for an individual.”
Privacy in Deep Learning: A Survey (2020, methodology): “This method was used for protecting patients while processing medical data and has been shown to fail on many occasions [49, 19, 50].”
Exploiting Unintended Feature Leakage in Collaborative Learning (2018, background): “Membership inference attacks against aggregate statistics are well-known [14, 27, 50], and recent work demonstrated black-box membership inference against ML models [24, 34, 53, 58], as discussed in Section 10.”
Tight Lower Bounds for Differentially Private Selection (2017, background): “Relationship to Previous Lower Bounds and Attacks Our argument is closely related to the work on tracing attacks [29], [30], [16], [17], [18], [26], [31].”
Comprehensive Privacy Analysis of Deep Learning: Stand-alone and Federated Learning under Passive and Active White-box Inference Attacks (2018, background): “In a tracing attack, however, the attacker’s objective is to learn if a particular individual’s data was included in the training dataset [4]–[6].”
Knock Knock, Who's There? Membership Inference on Aggregate Location Data (2017, neutral): “[15] show that one can learn whether a target individual was part of a case-study group associated to a certain disease by comparing the target’s profile against the aggregates of the case study and those of a reference population obtained from public sources.”
Privacy in Pharmacogenetics: An End-to-End Case Study of Personalized Warfarin Dosing (2014, neutral): “[20] and Sankararaman et al.”
Auditing Data Provenance in Text-Generation Models (2018, background): “Membership inference attacks against aggregate statistics have been demonstrated in the context of genomic studies [13], location time-series [26], and noisy statistics in general [8].”
Comprehensive Privacy Analysis of Deep Learning: Passive and Active White-box Inference Attacks against Centralized and Federated Learning (2018, background): “[4] performed one of the first membership inference attacks on genomic data.”
LOGAN: Membership Inference Attacks Against Generative Models (2017, background): “Additional membership inference attacks focus on genomic research studies [24, 5], whereby an attacker aims to infer the presence of a particular individual’s data within an aggregate genomic dataset, or aggregate locations [50].”
Robust Traceability from Trace Amounts (2015, background): “• Work in the genomics community [1], [2], [28], [29], [31] has so far focused on the case where exact statistics are available to the attacker (α = 0 in our formalism).”
Machine Learning with Membership Privacy using Adversarial Regularization (2018, background): “In particular, an adversary, with even black-box access to a model, can perform a membership inference [23, 43] (also known as the tracing [17]) attack against the model to determine whether or not a target data record is a member of its training set [45].”
LOGAN: Evaluating Privacy Leakage of Generative Models Using Generative Adversarial Networks (2017, background): “A few efforts [19, 3, 14] focus on inferring the presence of genomic data of particular individuals within a dataset of complex genomic mixtures.”