A multivariate analysis approach to the integration of proteomic and gene expression data

Area of Science:

• Systems Biology
• Bioinformatics
• Computational Biology

Background:

• Understanding cellular processes requires integrating diverse biomolecular data, including gene expression and proteomics.
• Current methods often rely on direct gene/protein correlations, which can be misleading due to complex regulatory mechanisms.
• Post-transcriptional and translational modifications complicate the direct relationship between gene expression and protein abundance.

Purpose of the Study:

• To apply a multivariate statistical method, co-inertia analysis (CIA), for integrating gene expression and proteomic data.
• To visualize relationships between gene expression and protein abundance data from the same biological samples.
• To project gene ontology (GO) information onto these visualizations to identify biologically relevant cellular processes.

Main Methods:

• Utilized co-inertia analysis (CIA), a multivariate statistical technique, to analyze and integrate multiple biological datasets.
• Employed principal components analysis (PCA) and correspondence analysis (CA) for single-dataset exploration.
• Integrated gene ontology (GO) annotations to interpret the biological meaning of identified patterns.

Main Results:

• Successfully visualized gene expression, protein abundance, and GO classes in low-dimensional projections.
• Applied the methodology to human malarial parasite and NCI-60 cancer cell line datasets.
• Identified specific GO classes likely to be of significant biological importance within the analyzed datasets.

Conclusions:

• Co-inertia analysis (CIA) provides a powerful framework for integrating and visualizing multi-omics data.
• This approach enhances the understanding of complex biological systems by linking gene expression and protein abundance.
• The integration of GO terms aids in the functional interpretation of molecular data, highlighting key cellular processes.