Effect of Tryptic Digestion on Sensitivity and Specificity in MALDI-TOF-Based Molecular Diagnostics through Machine...

Area of Science:

• Biochemistry and Analytical Chemistry
• Biotechnology and Bioinformatics

Background:

• Protein analysis using MALDI-TOF mass spectrometry (MS) can be limited by sample complexity.
• Machine learning (ML) algorithms applied to MALDI-TOF spectra can aid in bacterial sorting and disease diagnosis.
• Proteolytic digestion is a common sample preparation technique to simplify protein analysis.

Purpose of the Study:

• To evaluate the impact of proteolytic digestion on the sensitivity and specificity of MALDI-TOF MS coupled with ML for dichotomous sample sorting.
• To determine if protein digestion enhances biomarker discovery and diagnostic accuracy.

Main Methods:

• Four case studies were analyzed using MALDI-TOF MS in both intact protein mode and after overnight tryptic digestion.
• Machine learning software classified samples into dichotomous groups based on spectral data.
• Performance metrics including sensitivity, specificity, and Youden index were calculated to assess ML model effectiveness.

Main Results:

• Proteolytic digestion significantly improved the sensitivity and specificity of ML-based sorting for most samples.
• In cases with substantial chemical differences, intact and digested protocols yielded similar performance.
• The study demonstrated enhanced pattern recognition and improved diagnostic potential with digested samples.

Conclusions:

• Proteolytic digestion prior to MALDI-TOF MS analysis can substantially enhance the performance of ML-based classification and biomarker discovery.
• Digestion is not essential when samples are already easily distinguishable by their chemical composition.
• This workflow optimization holds promise for improved diagnostic tools and biological sample analysis.