Lung cancer prediction using multi-gene genetic programming by selecting automatic features from amino acid...

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Lung cancer is a leading cause of cancer-related mortality worldwide.
• Early diagnosis remains challenging due to reliance on late-stage physical features and complex somatic mutations.
• Identifying novel patterns in mutated genes and protein sequences is crucial for early cancer prediction.

Purpose of the Study:

• To develop an efficient predictive model for early-stage lung cancer detection.
• To analyze lung cancer-related mutated genes and protein amino acid sequences for novel predictive patterns.
• To implement an evolutionary learning technique for automatic feature selection and optimal combination.

Main Methods:

• Development of a biologically inspired multi-gene genetic programming algorithm.
• Utilizing discriminant information from protein amino acids for feature analysis.
• Efficiently selecting 23 discriminant features from an initial set of 1500 features.

Main Results:

• The proposed model achieved an Area Under the Receiver Operating Characteristic Curve (ROC AUC) of 98.79%.
• The system demonstrated a high accuracy of 95.67% in lung cancer prediction.
• The model outperformed existing lung cancer prediction approaches in performance metrics.

Conclusions:

• The developed evolutionary learning technique provides an efficient method for early lung cancer prediction.
• The model aids in understanding the complex and heterogeneous nature of lung cancer.
• This approach highlights the potential of analyzing genetic and protein sequence data for improved cancer diagnostics.