EXIMS: an improved data analysis pipeline based on a new peak picking method for EXploring Imaging Mass Spectrometry...

Area of Science:

• Biotechnology
• Bioinformatics
• Analytical Chemistry

Background:

• Matrix Assisted Laser Desorption Ionization-Imaging Mass Spectrometry (MALDI-IMS) generates high-volume 'omics' data with spatial information.
• Existing MALDI-IMS data processing methods often neglect spatial data, limiting the discovery of molecular distribution patterns.
• A novel pipeline is needed to fully exploit the spatial dimension inherent in MALDI-IMS.

Purpose of the Study:

• To develop and present a streamlined, unsupervised data analysis pipeline for MALDI-IMS data.
• To effectively utilize spatial information for identifying hidden molecular distribution patterns.
• To improve the analysis of complex MALDI-IMS datasets.

Main Methods:

• The pipeline employs Sliding Window Normalization (SWN) and a novel spatial distribution-based peak picking method using Gray level Co-Occurrence (GCO) matrices.
• Gist descriptors and improved GCO matrices are used for feature extraction from molecular images.
• Clustering of biomolecules and minimum medoid distance are applied for group identification and automatic group number estimation.

Main Results:

• The algorithm successfully revealed significant molecular distribution patterns in a Eucalypt leaf metabolomics dataset, which were missed by conventional methods.
• The spatial peak picking method outperformed traditional approaches on a rat brain proteomics dataset.
• While SWN showed no significant improvement over no normalization, visual assessment indicated benefits over median normalization.

Conclusions:

• The proposed MALDI-IMS data analysis pipeline effectively utilizes spatial information to uncover complex molecular distribution patterns.
• This novel approach enhances the analytical capabilities of MALDI-IMS for 'omics' studies.
• The freely available source code and data facilitate further research and application.