基于马尔可夫链的基因芯片特异性探针选择方法

对于基因表达芯片,特异性探针的选择是探针设计的重要环节,由于基因组序列数据量极大,不可能对每个候选探针都在全序列中进行特异性评价并进行取舍。对此问题,提出了一种采用马尔可夫链概率准则的探针特异性选择方法,即把基因组序列看作马尔可夫链,任何探针序列的互补序列作为它的一个子序列,都具有一定的出现概率,概率越小,越可能具有特异性。据此,选择其中概率最小的N个候选探针,能够大大减少进行特异性评价的探针数量,缩短探针设计的计算时间。对实际数据的测试结果表明,该方法选择的探针具有很高的特异性。

Method of selecting special probes for microarray based on Markov chain

Selecting special probes is an important step for expression arrays during probe designing. Because of the huge data in genome, it is impossible to judge the specialization of all possible probes in genome. In this paper, we developed a new method based on Markov chain to select special probes for microarray. Because genome sequence can be looked as a Markov chain, each probe, as a complemented subsequence of the genome, has some appearance possibility. The smaller the possibility is, the more likely the sequence would be a special one. So we can select the smallest N sequences as candidate probes to reduce the number of probes which should be evaluated by searching in the genome, and to shorten the probe designing time. Applying this method to the real data, the result shows that the probes selected by this method are almost special.