Protocol for Classification Single-Cell PBMC Types from Pathological Samples Using Supervised Machine Learning.

Area of Science:

• Immunology
• Computational Biology
• Genomics

Background:

• Peripheral blood mononuclear cells (PBMC) are crucial for studying the immune system, diseases, and vaccines.
• Single-cell transcriptomics (SCT) enables cell type identification via gene expression but often relies on complex manual classification methods.
• Accurate cell type classification is vital for disease diagnosis and evaluating treatment efficacy.

Purpose of the Study:

• To present a protocol for classifying PBMC cell types using supervised machine learning (ML) on SCT data.
• To enable high-accuracy and efficient classification of PBMC from pathological samples.
• To provide a framework applicable to various SCT platforms, including 10× Genomics.

Main Methods:

• The protocol involves three key stages: data preprocessing, training supervised ML models on labeled PBMC SCT datasets, and applying these models to new disease samples.
• Utilizes supervised machine learning algorithms for cell type classification.
• Focuses on single-cell transcriptomics data analysis.

Main Results:

• The developed protocol achieves high accuracy and efficiency in classifying PBMC cell types.
• Demonstrates the effectiveness of supervised ML for analyzing complex SCT data.
• The method is adaptable to different SCT platforms.

Conclusions:

• Supervised machine learning offers a powerful and efficient approach for PBMC classification from SCT data.
• This protocol facilitates more precise disease classification and therapeutic assessment.
• The methodology is broadly applicable to SCT datasets from various sources and technologies.