A simple strategy for sample annotation error detection in cytometry datasets.

Area of Science:

• Immunology
• Biotechnology
• Clinical Research

Background:

• Sample mislabeling in clinical studies compromises data integrity and can lead to erroneous conclusions.
• While quality control for sample identification is standard in genomics, it is not common practice for cytometric data.
• Accurate sample tracking is crucial for longitudinal and cross-sectional studies involving complex biological data.

Purpose of the Study:

• To develop and validate a method for detecting sample identification errors in cytometric datasets.
• To leverage human leukocyte antigen (HLA) class I allele expression as a biomarker for sample verification.
• To enhance the reliability and quality control of cytometry-based clinical research.

Main Methods:

• Utilized a 33-marker CyTOF (Cytometry by Time-Of-Flight) panel to measure HLA-A*02 and HLA-B*07 expression.
• Analyzed three longitudinal samples from 41 participants, assessing immune cell types and HLA allele expression.
• Cross-validated discordant samples using quantitative polymerase chain reaction (qPCR) for HLA class I allele data.

Main Results:

• Identified 3 out of 123 samples (2.4%) with HLA allele expression inconsistent with their longitudinal counterparts.
• These discrepant samples also showed a mismatch in their cytometric signature compared to qPCR HLA data.
• The findings confirm the utility of HLA expression profiling for identifying sample labeling errors.

Conclusions:

• The presented HLA allele expression method is effective for detecting sample mislabeling in longitudinal cytometric data.
• This technique can be integrated with other methods like GWAS or PCR for error detection in cross-sectional studies.
• Widespread adoption of this quality control strategy will significantly improve the integrity of clinical studies using cytometry.