Colon cancer prediction with genetic profiles using intelligent techniques

Area of Science:

• Bioinformatics
• Computational Biology
• Cancer Genomics

Background:

• Gene expression profiles offer insights into cellular states and disease mechanisms.
• Accurate tumor classification is crucial for effective cancer treatment strategies.

Purpose of the Study:

• To evaluate the efficacy of different machine learning classifiers for colon cancer classification using gene expression data.
• To identify key genes indicative of colon cancer through feature selection.

Main Methods:

• Utilized a benchmark colon cancer dataset for gene expression analysis.
• Employed t-statistic for feature selection to identify top-ranking genes.
• Compared the performance of Support Vector Machine (SVM), Neural Networks (Multi-Layer Perceptron - MLP), and Logistic Regression for class prediction.

Main Results:

• Support Vector Machine (SVM) achieved the highest classification accuracy, as indicated by the Area Under the ROC Curve (AUC) and overall accuracy.
• Logistic Regression performed as the second-best classifier, followed by Multi-Layer Perceptron (MLP).
• The top 10 selected genes are recognized for their differential expression in colon cancer.

Conclusions:

• SVM combined with t-statistic feature selection provides an efficient and reliable method for colon cancer classification.
• This approach offers a viable alternative to existing techniques for analyzing gene expression data in cancer research.