Fractal and Machine Learning Analyses of MALDI-TOF Mass Spectrometry Data in Glioblastoma

Area of Science:

• Proteomics
• Computational Biology
• Machine Learning

Background:

• Matrix-assisted laser desorption/ionization-time-of-flight mass spectrometry (MALDI-TOF MS) is crucial for disease biomarker discovery.
• Traditional MALDI-TOF MS data analysis relies on protein/peptide peak extraction for machine learning.
• Novel feature extraction methods are needed to enhance diagnostic capabilities.

Purpose of the Study:

• To investigate fractal analysis as a novel feature extraction technique for MALDI-TOF MS data.
• To assess the efficacy of fractal dimensions in differentiating glioblastoma patients from controls.
• To explore the potential of fractal features to improve machine learning model performance.

Main Methods:

• MALDI-TOF spectra were treated as time-series data.
• Fractal dimensions were calculated for each spectrum using various algorithms.
• Machine learning models were trained using fractal dimensions for glioblastoma diagnosis.

Main Results:

• Fractal dimensions alone achieved accurate glioblastoma diagnosis models.
• Fractal analysis underperformed compared to traditional feature extraction methods.
• Incorporating fractal features improved the performance of machine learning models.

Conclusions:

• Fractal analysis offers a new computational approach for MALDI-TOF MS data.
• Fractal dimensions show potential for disease diagnosis, including glioblastoma.
• This method expands the toolkit for mass spectrometry data analysis in proteomics.