快速搜索相似DNA子序列的抗失真哈希算法
Distortion-Resistant Hashing for rapid search of similar DNA subsequence
摘要：
生物信息学的基本任务之一是定位一个短的子序列$S$，在测序时读取，在一个长的参考序列$R$，如人类基因组。一种自然的快速方法是找到$S$的散列值，并将其与$R$的每个长度$|S|$序列的散列值的准备数据库进行比较。这种方法的问题在于，它只会发现一个完美的匹配，而实际上有很多小的变化：替换，删除和插入。这个问题可以修复，如果有一个哈希函数设计为容忍一些小的失真，相应的比对度量（如Needleman-Wunch）：设计为使两个相似的序列应该最有可能给出相同的哈希值。本文讨论了抗失真哈希（DRH）的建设，以产生这样的指纹快速搜索相似的序列。所提出的方法是基于率失真理论：在一个几乎均匀的长度$|S|$序列的子集，哈希值表示最接近的序列$S$。这提供了对冲突距离的一些控制：具有相同哈希值的序列。
英文摘要：
One of the basic tasks in bioinformatics is localizing a short subsequence $S$, read while sequencing, in a long reference sequence $R$, like the human geneome. A natural rapid approach would be finding a hash value for $S$ and compare it with a prepared database of hash values for each of length $|S|$ subsequences of $R$. The problem with such approach is that it would only spot a perfect match, while in reality there are lots of small changes: substitutions, deletions and insertions. This issue could be repaired if having a hash function designed to tolerate some small distortion accordingly to an alignment metric (like Needleman-Wunch): designed to make that two similar sequences should most likely give the same hash value. This paper discusses construction of Distortion-Resistant Hashing (DRH) to generate such fingerprints for rapid search of similar subsequences. The proposed approach is based on the rate distortion theory: in a nearly uniform subset of length $|S|$ sequences, the hash value represents the closest sequence to $S$. This gives some control of the distance of collisions: sequences having the same hash value.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
