Interdependent Genomic Privacy in Biological Networks

You are reading version 1 of this direction, frozen 2026-10-04. The latest version is version 2.

Version 1 of 221 papers40 conceptsfrozen 2026-10-04

In traditional data privacy, we assume that an individual has the sole authority to disclose their own information. If you choose to keep your social security number or web browsing history private, that choice is generally effective. However, genomic data breaks this model. Because DNA is shared across biological relatives, your genetic privacy is not a solo endeavor—it is a collective resource.

Research into interdependent genomic privacy explores how the disclosure of one person’s genetic markers can compromise the sensitive information of their non-consenting relatives.

The Biological Linkage as a Privacy Leak

The fundamental mechanism behind this research is the predictable nature of genetic inheritance. Humans share approximately 50% of their DNA with their parents and siblings, 25% with grandparents, and 12.5% with first cousins. Jun Sakuma argues in Genome and Privacy that this makes the genome the "ultimate personal information"—it is immutable, highly identifiable, and inherently shared.

When an individual uploads their data to a public or commercial database, they aren't just revealing their own predispositions; they are providing a partial map of their entire family tree. Erman Ayday, Jean‐Pierre Hubaux, and Amalio Telenti describe this in Interdependent Privacy Games: The Case of Genomics as a negative externality: the benefits of genetic testing (like health insights) go to the individual, but the privacy risks are distributed across their kin.

Quantifying Risk through Graphical Models

To understand how much information is leaked, researchers use probabilistic graphical models. By treating a family pedigree as a network of nodes, researchers like Jacques Fellay and Mathias Humbert apply belief propagation algorithms to estimate the genotypes of unobserved family members.

In Practical Solutions for Protecting Individual Genomic Privacy, the authors demonstrate that if an attacker knows the genetic data of a few relatives and the structure of the family tree, they can reconstruct the private genetic profile of a target individual with high statistical confidence. This "reconstruction attack" leverages Mendel’s laws and linkage disequilibrium—the tendency for certain DNA sequences to be inherited together—to fill in the blanks of a non-participant's genome.

The 2% Tipping Point

The scale of this vulnerability is non-linear. You do not need a majority of the population to share their DNA for everyone to lose their anonymity. Yaniv Erlich, Tal Shor, and Shai Carmi demonstrated in Re-identification of genomic data using long range familial searches that once a database covers roughly 2% of a population, nearly any individual can be matched to at least a third cousin.

This creates a "tipping point" in genomic privacy. Because third-cousin matches can be combined with basic demographic data (like age or location) to uniquely identify a person, the decision of a small minority to participate in services like GEDMatch effectively de-anonymizes the entire group. Ardalan Naseri and Kecong Tang further show in Personalized genealogical history of UK individuals inferred from biobank-scale IBD segments that large-scale biobanks allow for the reconstruction of up to 80% of an individual's genome using shared "Identity-By-Descent" (IBD) segments from distant relatives.

The Tragedy of the Genomic Commons

Why do people continue to share despite these risks? Researchers apply game theory to model this behavior. In When to Share Your Genome to an Online Community, Jia Guo and Murat Kantarcioglu describe a "tragedy of the commons."

As more relatives share their data, the marginal privacy remaining for a non-participant decreases. Eventually, a person may decide to join a database simply because their privacy is already effectively lost; they might as well gain the medical or genealogical benefits of the data since the "cost" (privacy loss) has already been paid by their cousins. This leads to a stable equilibrium where mass disclosure becomes the norm, regardless of individual preferences.

Moving Toward Collective Protection

Recognizing that individual consent is insufficient, the field is shifting toward technical and communicative safeguards:

  1. Privacy-Enhancing Technologies (PETs): Hyunghoon Cho and others advocate for Secure Multi-party Computation (SMPC) and Homomorphic Encryption, which allow researchers to compute genetic risks without ever seeing the raw DNA sequences.
  2. Quantifying Leakage: Erman Ayday proposes treating genomic privacy as a "depleting resource" in Towards Quantifying and Preventing the Leakage of Genomic Data, where every query against a database must be weighed against the total remaining privacy budget of the individual and their kin.
  3. User Awareness: Tools like the KGP Meter, developed by Didier Dupertuis and Kévin Huguenin, attempt to bridge the cognitive gap. In KGP Meter: Communicating Kin Genomic Privacy to the Masses, they show that while people often ignore altruistic risks to relatives, they are more likely to reconsider sharing when shown how their relatives' data specifically compromises their own personal health privacy.

Papers in this version

History & citationVersion 2 · 2026-10-04

Current · Permanent link

  1. Version 2 (latest) · 2026-10-04 · 18 papers
  2. Version 1 · 2026-10-04 · 21 papers
Cite this direction

Latest version (a living page):

@misc{lacuna_interdependent_privacy_c2659eba,
  title = {interdependent privacy},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/interdependent-privacy/rdr_c2659eba77094a4db0ecd43fd2bfb0a7},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-19},
}

This exact version (version 1, frozen text):

@misc{lacuna_interdependent_genomic_privacy_in_biolog_c2659eba_v1,
  title = {Interdependent Genomic Privacy in Biological Networks},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/interdependent-genomic-privacy-in-biological-networks/rdr_c2659eba77094a4db0ecd43fd2bfb0a7/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-19},
}