Genomic data privacy is an essential thing while sharing the genomic data to the public. How can the privacy of genomic data be protected? Which anonymization models are useless for preserving the privacy of genomic data? Which model is suggested for preserving the privacy?
I think "useless" might be too strong of a word, often the limitations of anonymization are appropriately disclosed in the patient consent. But a recent paper in Science showed that using only genome sequence and the state of residence of the person providing the DNA sample together with public databases of surname frequency (in the U.S.), one could re-identify many sample contributors. The paper was by Gymrek, Science vol 339 no. 6117 p. 321-324, "Identifying personal genomes by surname inference". There are some interesting techniques that involve replacement of valid sequence with randomly introduced variants that are tracked with a key that is in the hands of the investigator or data manager. Such modified sequence could not be publicly shared without clearly disclosing that the data is modified.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics