Mechanisms for more efficient antibiotics and antibiotic resistance genes removal during industrialized treatment of over 200 tons of tylosin and spectinomycin mycelial dregs by integrated meta-omics.

To mitigate the environmental risks posed by the accumulation of antibiotic mycelial dregs (AMDs), this study first attempted over 200 tons of mass production fermentation (MP) using tylosin and spectinomycin mycelial dregs alongside pilot-scale fermentation (PS) for comparison, utilizing the integrated-omics and qPCR approaches. Co-fermentation results showed that both antibiotics were effectively removed in all treatments, with an average removal rate of 92%. Antibiotic resistance gene (ARG)-related metabolic pathways showed that rapid degradation of antibiotics was associated with enzymes that inactivate macrolides and aminoglycosides (e.g., K06979, K07027, K05593). Interestingly, MP fermentations with optimized conditions had more efficient ARGs removal because homogenization permitted faster microbial succession, with more stable removal of antibiotic resistant bacteria and mobile genetic elements. Moreover, Bacillus reached 75% and secreted antioxidant enzymes that might inhibit horizontal gene transfer of ARGs. The findings confirmed the advantages of MP fermentation and provided a scientific basis for other AMDs.

Manifold Interpolation for Large-Scale Multiobjective Optimization via Generative Adversarial Networks.

Large-scale multiobjective optimization problems (LSMOPs) are characterized as optimization problems involving hundreds or even thousands of decision variables and multiple conflicting objectives. To solve LSMOPs, some algorithms designed a variety of strategies to track Pareto-optimal solutions (POSs) by assuming that the distribution of POSs follows a low-dimensional manifold. However, traditional genetic operators for solving LSMOPs have some deficiencies in dealing with the manifold, which often results in poor diversity, local optima, and inefficient searches. In this work, a generative adversarial network (GAN)-based manifold interpolation framework is proposed to learn the manifold and generate high-quality solutions on the manifold, thereby improving the optimization performance of evolutionary algorithms. We compare the proposed approach with several state-of-the-art algorithms on various large-scale multiobjective benchmark functions. The experimental results demonstrate that significant improvements have been achieved by the proposed framework in solving LSMOPs.