Scientists develop accurate medical analysis tools for cancer, iCAGES

With the increasing number of accurate medical-related analytical methods for cancer, the continuous development of new analytical tools will not only be limited to the analysis method itself, but also focus on improving the user interaction experience, "one-stop" and other comprehensive services, targeting the patient's specific situation. The specific analysis aims to provide the most comprehensive and accurate analysis service without requiring users to have any ability to interpret biological information.

Recently, researchers from the Columbia University Medical Center have developed a set of precision treatments, iCAGES, that can quickly compare and screen related cancer genes and search for and suggest these genes. This is not only the most comprehensive information available in the same type of analysis tool, but also the first analysis tool to provide a clean user interface.

Most cancers are derived from the accumulation of somatic mutations, not a single gene mutation. On the other hand, large-scale sequencing projects for cancer patients to date have found that many mutations play a role in the development of many different cancers. However, these mutational information does not provide any specific information on cancer screening at the individual level: mutations that actually cause cancer may differ from each other, and some mutations thought to be involved may have no significant effect on a person. At present, no clinically used analytical tool can predict cancer-related mutations at the individual level. “Even even if the doctor knows the genes that cause cancer in the patient, it is not easy to choose the most suitable one from the hundreds of possible drug therapies,” said Dr. Wang Kai, the development leader of the tool. Speaking of it.

To fill this gap, they developed the latest analytical tool, the Cancer Genome Scoring Integration Tool (iCAGES). First, it analyzes the patient's genome and compares it with the sequencing information of the tumor tissue to narrow the screening range to determine possible induced mutations. Next, iCAGES compares the mutation to the existing cancer mutation database, completes the corresponding annotations, and combines statistical analysis methods such as machine learning to screen out the most likely cancer-causing genes in this patient. Finally, based on the type of mutations screened, search for all targeted therapies that have been approved by the FDA. The entire process takes 30 minutes. In contrast, traditional methods often have many steps that require manual input, often taking weeks.

In a test using real patient data, iCAGES used the sequencing data of a lung cancer patient to screen out the most likely gene ARAF from the known 129 induced genes and recommended it from 122 different drug therapies. Sorafenib is the best treatment. This is consistent with the conclusions of the patient's attending physician, but using a longer-lasting traditional treatment. It is worth mentioning that sorafenib is not primarily a drug for lung cancer, which also reveals the possibility that iCAGES may give new treatments for existing drugs.