Base on artificial neural networks, people can observe an impressive amount of gene data expression, also discover specific groups of disease-related genes. Researchers at Linköping University had published this important study in Nature Communications that they all looking for whether the method can be applied within accurate medicine and personalized treatments.
For any social media platforms that you commonly use, the suggestions feature may recommends you whom you may want to add contact with. This process depends on you and others’ activities that may have common contact, which indicates that you might have known each other before. Similarly, the maps that scientists trying to create based on the interaction between genes and proteins. The new trial using artificial intelligence, or AI, to investigates the possibility of adding deep learning as a method to discover biological networks, entities known as “artificial neural networks” being trained by experimental data. Since this networks get used to learn how to search for patterns in complex data storage, it is easy to transform this into applications such as image recognition. Unfortunately, this method hasn’t been popularized in biological researches yet.
The scientists contain any databases with information about the expression patterns of over 20,000 genes correspond with the same number of samples. The “raw” information, in terms of researchers did not attach the network information about which gene belongs to healthy or people who have diseases.
There is one of the challenges for machine learning: Is it possible to get exactly how an artificial neural network can solve a given task. AI sometimes was called as a “black box” – we only get what we have put into the box and it produced results. No specific steps included. The networks consist an input and out layer that bring the result of the carry out information processing by the whole system. When the network been trained, scientists wondered whether they could “lift” the lid of the black box and understand the process.
The scientists then investigated to confirm that whether the model of gene expression could be used to determine which patterns are abnormal and which is fine. The result showed that the model find relevant which agree well with biological mechanisms in the body. Since it has been trained using unclassified data, it proves that the network has possibly found brand new patterns, relevant form a biological respective.
About us:
TMA Solutions was established in 1997 to provide quality software outsourcing services to leading companies worldwide. We are one of the largest software outsourcing companies in Vietnam with 2,600 engineers. Our engineering team was selected from a large pool of Vietnam IT resources; they are well-trained and have successfully completed many large and complex projects. Please visit us: https://www.tmasolutions.com/
No comments:
Post a Comment