多目标优化的非编码RNA比对及预测

Alignment and Prediction of Multi-objective Optimized Non-coding RNA

为更好地优化多个代价函数,提出一种多目标模拟退化算法,在运算过程中对多目标进行优化,从而得到边界上不同方向的最优解,介绍进化过程中的非编码RNA结构,并在此基础上提出RNA多重比对及预测并行模型。实验结果表明,该模型能有效提高解的精度和多样性。

To optimize multiple cost functions better, a multi-objective simulated annealing algorithm is proposed. In the process of operation, the multiple objectives are optimized. The optimal solutions of different directions on boundary are obtained. The structure of non-coding RNA is introduced in process of evolution. On the basis of this, the multiple comparison and prediction parallel model is presented. Experimental results show this model can promote the accuaracy and diversity of the solutions.