Prognosis Analysis for Ovarian Cancer Patients Using Protein Data

Ovarian cancer is one of the most common genetic diseases caused by genetic mutations or chromosomal abnormal-ities. In the research field, many researchers have made efforts to identify cancer biomarkers using various genomic approaches. However, genomic changes are not the only factors that determine the phenotype of cancer cells, as millions of various factors are expressed in ovarian cancer. Therefore, there is a problem that research on identifying symptoms in cancer patients based on data is insufficient. To solve this problem, we propose a key factor selection technique based on unsupervised learning as well as a novel data preprocessing technique suitable for medical data with a large number of features. By applying this method, it is possible to select protein factors that could diagnose the patient's prognosis. Also, the effectiveness of the proposed technique is proven based on the TMA cohort process. With the proposed method, it is possible to indirectly analyze the prognosis of cancer patients and apply customized treatment methods for each patient. Furthermore, we could implement solutions and products that could periodically monitor the patient's condition.