Using Classification and K-means Methods to Predict Breast Cancer Recurrence in Gene Expression Data.

Area of Science:

• Bioinformatics
• Genomics
• Cancer Research

Background:

• Breast cancer affects approximately 10% of women.
• Early prediction of recurrence is crucial for patient outcomes.
• Gene expression data offers insights into cancer biology.

Purpose of the Study:

• To develop and evaluate a new method for predicting breast cancer recurrence.
• To identify key genes associated with cancer recurrence.
• To compare the performance of various machine learning techniques.

Main Methods:

• A novel method involving data collection, clustering, and classification was applied.
• Eight distinct techniques, including random forest and support vector machines, were implemented.
• The study analyzed 12,172 genes across 200 samples.

Main Results:

• Thirty differentiating genes were identified for classification.
• The random forest + k-means technique demonstrated superior performance.
• Neural network + k-means and random forest + k-means excelled in identifying high-risk cases.

Conclusions:

• Clustering significantly improved the performance of classification techniques.
• The developed method effectively utilizes gene expression data for recurrence prediction.
• Identifying high-risk breast cancer patients can be enhanced through advanced computational methods.