单基因扰动实验的自适应灰关联聚类算法

摘要

为克服单基因扰动实验必须遍历所有实验基因所导致的周期长和成本高的缺点,提出了一个自适应参数的灰关联聚类算法,每次基因扰动实验后通过自适应参数灰关联聚类算法选择下一个实验基因,从而省略不必要的基因扰动实验,得到满意的基因扰动顺序并建立基因调控网络.实验结果表明,该算法能有效地甄别对构建基因调控网络基本没有影响的基因,算法结果令人满意.%The disadvantages of single-gene disturbance experiment which is caused by traversing all genes of the experiment are the longer cycle and the higher cost. An auto-selected parameter grey relation cluster analysis method is developed. By applying it to the gene experiment data, the method can effectively identify those genes which have little influence to construct a satisfactory gene regulation network, the result of the calculation experiment is satisfied.