A robust biomarker discovery pipeline for high-performance mass spectrometry data

Area of Science:

• Biochemistry
• Computational Biology
• Oncology

Background:

• Accurate biomarker determination is crucial for early disease detection.
• Traditional mass spectral data analysis can be confounded by complex peak shapes and overlapping signals.
• High-performance mass spectrometry generates complex datasets requiring sophisticated analytical tools.

Purpose of the Study:

• To develop a high-throughput software pipeline for analyzing high-performance mass spectral data.
• To enable rapid and accurate biomarker determination, particularly for cancer detection.
• To bypass limitations of traditional peak-finding algorithms in mass spectrometry.

Main Methods:

• Developed a software pipeline utilizing discrete m/z data points, bypassing traditional peak-finding steps.
• Implemented methods to assess data set quality and suitability of m/z values as biomarkers.
• Applied the algorithm to serum mass spectra from ovarian cancer patients and healthy controls.

Main Results:

• Identified and ranked potential biomarker candidates based on their ability to discriminate between cancer and non-cancer conditions.
• Achieved a classification sensitivity of 95.6% and specificity of 97.1% using a simple distance calculation.
• Demonstrated superior performance compared to the CA125 marker for ovarian cancer detection.

Conclusions:

• The developed software pipeline effectively extracts biomarker candidates from high-performance mass spectral data.
• The novel approach offers improved sensitivity and specificity for disease biomarker discovery.
• This analytical package has the potential to enhance early cancer diagnosis and monitoring.