Understanding Robust Adaptation Dynamics of Gene Regulatory Network.

Robust adaptation is a critical attribute for gene regulatory network (GRN), understanding the relationship between adaptation and the GRN topology, and corresponding parameters is a challenging issue. The work in this paper includes: first, seven constraint multiobjective optimization algorithms are used to find sufficient solutions to get more reliable statistic rules. Meanwhile, the algorithms are compared to facilitate the future algorithm selection; second, a fuzzy c-mean algorithm is used to analyze solutions and to classify the solutions into different groups; third, the histogram analysis for all satisfactory solutions shows the preferred parameter range, i.e., parameter motif. The contributions of this paper includes: 1) Two new adaptation indices i.e., peak time and settle down time, are proposed for the first time to give more accurate description of the robust adaptation. Our conclusion is that some solutions even with satisfactory sensitivity and precision are not practically of robust adaptation because of too long time needed. 2) The relationship between topology, parameter set, and robust adaptation of GRN is discovered in the sense of both preferred topology and parameter motif. Our conclusion is that the robust adaptation depends more on the GRN topology than the model parameter set in two feasible topologies.