Systems metabolic engineering (SME) integrates systems biology, synthetic biology, and evolutionary engineering to design high-performance tailor-made microbial cells. SME advances environmental biotechnology by delivering innovative, scalable, and sustainable solutions for green bioprocesses, waste valorization, pollutant degradation, and resource management. Recent advances in artificial intelligence and machine learning, dynamic metabolic control systems, and metabolite-responsive regulatory elements have further accelerated strain development and expanded the functional scope of engineered microbes and microbial consortia. Here, we discuss key conceptual foundations, technological advances, emerging research directions and current limitations in SME, highlighting its transformative potential to drive innovation toward environmentally resilient, resource-efficient, and sustainable biotechnological solutions.
Wet clutches are critical components in mechanical power transmission systems, enabling controlled torque transfer between rotating parts. However, when disengaged, viscous drag torque is generated, reducing the transmission efficiency. Surface texturing, particularly groove patterns, is commonly employed to modulate flow dynamics and aeration process, ultimately decreasing drag torque. Despite their practical importance, systematic approaches for optimising groove design in wet clutches remain limited in the literature. To address this gap, this study proposes an integrated experimental-Gaussian Process Regression-based, multi-objective optimisation methodology aiming to minimise cumulative power loss while maximising torque capacity. To achieve that, a single-disc test rig was designed to evaluate various groove geometries and operating conditions for the GPR model. Conventional radial grooves were benchmarked against new topologies, and exhaustive search optimisation was used to explore the design space and find the optimal solution. The findings demonstrate the potential for groove design optimisation, with quantified benefits such as up to 23% reduction in cumulative power loss for arc-bow grooves and up to 21.8% for mid-relief designs, highlighting their broader applicability in wet clutch and lubrication systems.
As generative AI advances, the distinction between authentic and synthetic media is increasingly blurred, challenging the integrity of online information. In this study, we present CONVEX, a large-scale dataset of multimodal misinformation involving miscaptioned, edited, and AI-generated visual content, comprising over 150K multimodal posts with associated notes and engagement metrics from X's Community Notes. We analyze how multimodal misinformation evolves in terms of virality, engagement, and consensus dynamics, with a focus on synthetic media. Our results show that while AI-generated content achieves disproportionate virality, its spread is driven primarily by passive engagement rather than active discourse. Despite slower initial reporting, AI-generated content reaches community consensus more quickly once flagged. Moreover, our evaluation of specialized detectors and vision-language models reveals a consistent decline in performance over time in distinguishing synthetic from authentic images as generative models evolve. These findings highlight the need for continuous monitoring and adaptive strategies in the rapidly evolving digital information environment.
We argue that learning visual representations without labels requires a training signal jointly complete across three non-overlapping objectives: semantic invariance across augmented views, patch-level spatial prediction, and representational non-degeneracy. We formalize these as the observation, prediction, and regularization principles and prove (i) that combining observation and prediction without regularization admits the constant encoder as a global minimizer under negative-free alignment; (ii) that the two objectives are gradient-complementary and structurally non-conflicting at the encoder output; and (iii) that the momentum encoder converges to the same fixed point as the online encoder and provides no collapse guarantee at convergence. Contrastive alignment provides only self-limiting collapse resistance, formalized via an explicit gradient-decay argument. Dropping prediction withholds the spatial training signal by construction; dropping observation forfeits cross-view semantic invariance by construction; at the scale we study, no pair substitutes for the third. Every major self-supervised method is a special case of a single unified energy decomposition. We pair every theoretical claim with a controlled experiment, including a patch-retrieval evaluation for the spatial consequence of prediction.
