Particle uptake in cancer cells can predict malignancy and drug resistance using machine learning

Area of Science:

• Biophysics
• Cancer Biology
• Machine Learning

Background:

• Tumor heterogeneity significantly impacts cancer treatment efficacy.
• Developing predictive tools for cancer cell classification is crucial for personalized therapy.
• Cellular mechanical properties are linked to cancer cell function and behavior.

Purpose of the Study:

• To classify cancer cell subpopulations using mechanical measurements via particle uptake.
• To investigate the utility of machine learning in analyzing single-cell particle uptake patterns.
• To assess the potential of this method for distinguishing drug-resistant or highly malignant cancer cells.

Main Methods:

• Utilized particle uptake assays with fluorescently labeled polystyrene particles (0.04–3.36 μm).
• Employed flow cytometry to analyze single-cell particle uptake patterns.
• Applied machine learning algorithms for classification of cancer cell subtypes.

Main Results:

• Machine learning algorithms achieved >95% accuracy in classifying cancer cell subtypes.
• Particle uptake patterns effectively differentiated morphologically similar cancer cell subpopulations.
• Uptake data demonstrated a 'normalization' effect, reducing experimental variation.

Conclusions:

• Particle uptake combined with machine learning offers a robust method for classifying cancer cells based on biomechanics.
• This technique shows promise for enhancing cancer therapy by enabling precise cell classification.
• The method provides a reliable way to identify and potentially target diverse cancer cell populations.