Classifying osteosarcoma patients using machine learning approaches.

Area of Science:

• Biochemistry
• Oncology
• Bioinformatics

Background:

• Osteosarcoma, a prevalent bone cancer, has seen limited progress with genomic and proteomic studies.
• Early detection of osteosarcoma is crucial for effective metastasectomy treatment.
• Metabolomic data analysis offers a novel approach to understanding and classifying osteosarcoma.

Purpose of the Study:

• To evaluate the efficacy of different machine learning models in classifying osteosarcoma using metabolomic data.
• To compare the performance of logistic regression, support vector machine (SVM), and random forest (RF) classifiers.
• To identify the most accurate method for early osteosarcoma detection.

Main Methods:

• Collected metabolomic data from osteosarcoma patients and healthy controls.
• Applied three classification algorithms: logistic regression, SVM, and RF.
• Utilized receiver operating characteristic (ROC) curves to evaluate classifier performance.

Main Results:

• All three classifiers successfully distinguished between osteosarcoma cases and healthy controls.
• Random forest (RF) demonstrated superior performance during cross-validation with 97% accuracy.
• The RF model achieved an overall accuracy of 95% and an AUC of 0.99 on the testing set.

Conclusions:

• Metabolomic data analysis, particularly using random forest, is a highly effective method for osteosarcoma classification.
• This approach holds significant potential for early detection and improved patient outcomes.
• Further research can leverage metabolomics for enhanced osteosarcoma diagnosis and treatment strategies.