Automated tissue analysis--a bioinformatics perspective.

Area of Science:

• Computational pathology
• Bioinformatics
• Digital pathology

Background:

• Automated tissue analysis (tissomics) enables reproducible phenotypical characterization of histological specimens.
• Synergies between tissue analysis and bioinformatics hold significant potential for medical research and computational diagnostics.

Purpose of the Study:

• To introduce informatics tools for clustering and correlating quantitative tissue profiles with gene expression data.
• To discuss the perspectives of combined tissue analysis and bioinformatics in medical research and computational diagnostics.

Main Methods:

• Review of key enablers in microscopic imaging and machine vision for high-throughput tissue analysis.
• Description of methodologies for combined analysis of tissue and gene expression profiles, emphasizing individual responses.

Main Results:

• Machine vision extracts comprehensive histomorphometric profiles, providing reproducible data on tissue components and heterogeneity.
• Quantitative tissue information, when integrated with bioinformatics data like gene expression profiles, allows for more comprehensive stratification of individual responses.
• Tissue data serve as co-variants in bioinformatics, aiding the identification of candidate genes relevant to tissue injury or disease.

Conclusions:

• Automated analytics generate quantitative tissue data, moving beyond traditional visual histopathology.
• Reproducible datasets from automated analysis can be statistically correlated and clustered with bioinformatics data.
• The combined approach facilitates a system-wide view of biology, potentially accelerating personalized computational diagnosis.