Advanced machine learning to crack the genetic code Stanford University developed a new tool to iden segotep

Advanced machine learning | crack the genetic code: Stanford University developed a new tool for the identification of pathogenic gene mutations (of paper) – Sohu technology from Standford Author: Erin Digitale machine of the heart: Cao Rui, Li Zenan participated in the compilation of 2012, Shayla Haddock’s doctor was tested on a rare genetic disease she had, but they no way to diagnose. Deformities, short stature, abnormal facial features, and congenital deafness will accompany her for life – which makes her doctors suspect that this is caused by a genetic mutation. But for children like Shayla, it is very difficult to find the causative genes in 3 billion DNA base pairs. After gene sequencing, almost every case requires a trained geneticist to spend 20-40 hours to analyze. And about 75% of patients could not be diagnosed in the first attempt. Some computer scientists Standford finally solved the dilemma of Shayla, they designed an automatic way, comparing the symptoms and gene mutation of patients with genetic diseases and the existing information in the database. In early 2016, they found that Shalya’s disease was reported earlier in the medical literature. And two weeks ago, Shayla’s doctors couldn’t tell her family what the cause was. Now, the Stanford University research team, in the computer scientists, geneticists, Dr. Gill, led by Dr. Bejerano, in the development of the diagnosis of the disease has taken a step forward. They developed a more sophisticated tool to automatically evaluate the errors in a single letter in the genetic code. This new tool is called M-CAP, their research papers have been published in today’s "Nature Genetics", the team used a machine learning algorithm, according to whether it is possible to classify the pathogenic gene mutation. Full details of the results have been published online for the use of genetic researchers around the world. "There are hundreds of thousands of changes in the human genome that can cause severe disease in children," he said. These variations are very different from those of healthy people. "He explained, the most fundamental problem is the synthesis of protein each person’s genetic code in part has 10 thousand small spot or a variant, each of these bases in DNA and normal human gene sequences of different. Almost all of these are harmless. But there are good reasons to believe that one or two genetic changes have caused the disease in children who are born with unexplained symptoms. In their assessment, geneticists have tried to one to two are most likely to cause the disease gene variant zero. For example, they would ignore the genetic variants that are common in the general population, because we believe that these rare diseases are caused by rare mutations that cause n相关的主题文章: