Classification of Tumor Metastasis Data by Using Quantum Kernel-Based Algorithms

Tumor metastasis is a dynamic process, and its fatality rate is relatively high. Different genes are regulated during the metastasis process. Next-generation sequencing technology is a new approach that enables rapid and high-throughput whole-transcriptome measurements. Support vector algorithms make use of kernel functions to classify data effectively. A few studies suggest that quantum support vector machine algorithms can perform well in classification problems. If biomarkers can be identified to predict tumor metastasis accurately, it will be an important step toward precision medicine. In this study, we use both the SVM and QSVM classifiers with the addition of a certain number of features, we can achieve very good distinctions between patients with or without metastasis. This is a positive result for precision medicine studies. Also, we evaluate the performance of quantum and classical algorithms in classifying tumor metastasis data. Our preliminary study indicates that the classical kernel-based classifier performs better than the quantum version.