Pan-cancer analysis for studying cancer stage using protein expression data

Area of Science:

• Oncology
• Proteomics
• Bioinformatics

Background:

• Pan-cancer analyses identify commonalities across multiple cancer types.
• Protein expression data offers direct insight into patient phenotypes.
• Understanding protein expression differences is crucial for cancer progression insights.

Purpose of the Study:

• To analyze differentially expressed (DE) proteins in early versus advanced cancer stages across various cancer types.
• To investigate the relevance of identified DE proteins using predictive modeling.
• To evaluate the utility of pan-cancer analysis for identifying biologically relevant proteins and improving cancer progression prediction.

Main Methods:

• Utilized reverse-phase protein array (RPPA) data for protein expression analysis.
• Performed differential expression analysis to identify DE proteins between early and advanced cancer stages.
• Developed predictive models using K-nearest neighbor (KNN) and linear discriminant analysis (LDA) classifiers.

Main Results:

• Identified specific DE proteins associated with cancer progression across multiple cancer types.
• Demonstrated the effectiveness of KNN and LDA classifiers in building predictive models based on protein expression.
• Highlighted the complementary value of pan-cancer analysis to single-cancer studies.

Conclusions:

• Pan-cancer protein expression analysis is valuable for discovering fundamental cancer patterns.
• This approach can identify biologically relevant DE proteins critical for cancer progression.
• Pan-cancer analysis aids in developing robust predictive models for cancer staging and patient outcomes.