British "Nature" magazine recently published a research result of cancer online. European scientists have developed a new machine learning program that can improve the diagnosis of brain tumors through DNA methylation data-"methylation fingerprinting" technology.
Correct diagnosis of the tumor is essential for the treatment of cancer. However, among the approximately 100 known tumors, central nervous system tumors are particularly difficult to accurately identify.
To solve this problem, Stefan Pfister, a scientist at the German Cancer Research Center, and colleagues developed a machine learning program that can classify methylation data. Methylation is the process by which methyl molecules attach to DNA and change the availability of information within the DNA. This process not only occurs naturally during the normal functioning of cells, giving each cell a unique "methylated fingerprint", it also occurs in diseases such as cancer. Therefore, methylation can reveal information about tumor types and “tell†people what type of cells these tumors form.
After the programs developed by the research team have been trained, 91 types of tumors can be identified using “methylated fingerprintsâ€. The reference data used in the training came from approximately 2800 cancer patients. Researchers conducted tests on 1,104 patients who had been manually examined for CNS tumors. They thought that the procedure not only improved the diagnostic accuracy, but also its objectivity made it possible to faithfully identify new types of rare tumors—manual examinations. There will be pressure to diagnose based on known tumor types, even in atypical cases.
In order to make this new method widely used, the research team earlier generated a free online tool that can analyze the uploaded data within a few minutes. Since its launch, the tool has been used more than 4,500 times, and users can choose to share their data in order to further optimize the algorithm. The researchers concluded that the integration of "methylated fingerprints" with brain tumor auto-classifiers can also provide a blueprint for creating similar tumor classification algorithms for diagnosing more other cancer types.
Striped Bonito,Sarda orientalis,Fresh Striped Bonito,Fresh Sarda orientalis
Zhoushan Boda Aquatic Products Co.,Ltd , https://www.baida-aquatic.com