Network indexing algorithm or will be applied to DNA search engine

As science and technology progressed, scientists decoded more and more genomes. But for geneticists or researchers, finding the shared genome of an organism is a difficult task-because there are too many DNA letters to compare. How to quickly search for a huge bioinformatics database. Most researchers use the blast or fastahttp://www.aliyun.com/zixun/aggregation/4793.html ">" algorithm, which essentially compares each pair of genomes. But now the third-largest search engine in China, a computer scientist at Tencent's search (soso.com), Liang, has put forward an event that uses Chinese indexing algorithms to retrieve biological information.

Liang pointed out that there is no gap between each word in Chinese, so one way to index Chinese documents is to decompose the text into n fragments (n-grams), n to represent words, 1-grams to represent a Chinese character, 2-grams to two characters, and 3-grams to represent three characters, Some Chinese search engines only index 2-grams. Liang said that the statistical distribution of DNA sequences should follow the law of Zipf (ZIPF). Zipf law refers to the frequency of the occurrence of a word in the frequency of the table and its ranking in inverse proportion, the highest frequency of the words appear frequency of the second word is about twice times the frequency, and the second occurrence of the frequency of the word is the fourth-place frequency of the word twice times. Liang used the same criteria to find the average length of the genomes of Arabidopsis, aspergillus, Drosophila and mice, and he found that the average length was 12 letters, so the 12-gram index of genomic data might be optimal.