A non-linear rectification based on immune genetic algorithm (IGA) is proposed in this paper, for the shortcoming of the non-linearity rectification. This algorithm introducing the biologic immune mechanism into the genetic algorithm can restrain the disadvantages that the poor precision, slow convergence speed and early maturity of the genetic algorithm. Computer simulations indicated that the algorithm not only keeps population diversity, but also increases the convergent speed, precision and the stability greatly. The results have shown the correctness and effectiveness of the method.
Algorithms ; Computer Simulation ; Immunity ; genetics

of Streptomyces regensis was isolated from soil to produce a novel antibiotic AGPM of a strong biological activity of antitumor The strain was irradiated by UV after treatment with LiCl to give a AGPM yield of 1 87?10 2 mg mL 1 , 2 2 times higher than that of the original strain The optimum UV irradiation time was 30～60 s and the best LiCl concentration was 0 05～0 09 mol/L The fermentation of AGPM was conducted in a 30 L stirred tank, the maximum yield of AGPM using the mutants reached 1 85?10 2 mg mL 1 , while that using the original strain was only 0 85?10 2 mg mL 1

A new strain of Streptomyces regensis was isolated from soil to produce a novel antibiotic AGPM possessing a strong antitumor activity In order to study on the metabolic path of the novel antibiotic AGPM, the protein patterns from the strain of Streptomyces regensis at different culture period were analyzed by using two eimensional polyacrylamide gel electrophoresis Comparing with sample from growth phase, seventeen new protein spots were found in that from antibiotic production phase The results demonstrated that the special proteins might be related with the antibiotic AGPM biosynthesis from Streptomyces regensis