Multidimensional profiling of cell surface proteins and nuclear markers.

Area of Science:

• Cellular Biology
• Computational Biology
• Biophysics

Background:

• Cell membrane proteins are crucial for tissue structure and intercellular communication.
• Understanding protein distribution at a single-cell level is key to identifying treatment effects.

Purpose of the Study:

• To develop and validate a computational method for segmenting and multidimensionally characterizing cell membrane proteins on a cell-by-cell basis.
• To improve the classification and visualization of different treatment groups by analyzing phenotypic signatures.

Main Methods:

• Developed computational steps for delineating cell membrane protein signals and associating them with nuclei.
• Computed coupled representations of DNA content and membrane proteins for multidimensional phenotypic analysis.
• Ranked features, visualized them using heatmaps, and discriminated between treatment groups.

Main Results:

• Demonstrated that multidimensional cell-by-cell phenotype representations enhance predictive and visualization capabilities.
• Successfully applied the method to cells treated with different radiation qualities and small molecules.
• Identified previously hidden variables influencing treatment group differences.

Conclusions:

• The novel method enables detailed, cell-specific analysis of membrane protein distribution and cellular phenotypes.
• This approach offers improved insights into cellular responses to various treatments, aiding in biomarker discovery.