On the privacy risks of sharing clinical proteomics data.

Area of Science:

• Biochemistry
• Bioinformatics
• Genomics

Background:

• Genomic study privacy is known, but clinical proteomic data privacy risks are understudied.
• Clinical proteomic datasets are increasingly used for patient analysis.

Purpose of the Study:

• To comprehensively analyze privacy risks in clinical proteomic data.
• To identify specific data points that pose privacy risks.

Main Methods:

• Analysis of clinical proteomic datasets from blood/serum samples.
• Identification of minor allelic peptides at non-synonymous single nucleotide polymorphism (nsSNP) sites.
• Assessment of patient identifiability from identified peptides.

Main Results:

• A small fraction of minor allelic peptides can identify individual patients with high confidence.
• Raw clinical proteomic data presents significant privacy risks.
• Removing identified minor allelic peptides effectively mitigates privacy risks.

Conclusions:

• Clinical proteomic data carries inherent privacy risks due to identifiable peptides.
• A simple data pre-processing step, removing minor allelic peptides, can significantly enhance data privacy.
• This mitigation strategy has minimal impact on the utility of proteomic datasets for further analysis.