Schizochytrium species are one of the best oleaginous thraustochytrids for high-yield production of docosahexaenoic acid (DHA, 22:6). However, the DHA yields from most wild-type (WT) strains of Schizochytrium are unsatisfactory for large-scale production. In this study, we applied the atmospheric and room-temperature plasma (ARTP) tool to obtain the mutant library of a previously isolated strain of Schizochytrium (i.e., PKU#Mn4). Two rounds of ARTP mutagenesis coupled with the acetyl-CoA carboxylase (ACCase) inhibitor (clethodim)-based screening yielded the mutant A78 that not only displayed better growth, glucose uptake and ACCase activity, but also increased (54.1%) DHA content than that of the WT strain. Subsequent optimization of medium components and supplementation improved the DHA content by 75.5 and 37.2%, respectively, compared with that of mutant A78 cultivated in the unoptimized medium. Interestingly, the ACCase activity of mutant A78 in a medium supplemented with biotin, citric acid or sodium citrate was significantly greater than that in a medium without supplementation. This study provides an effective bioengineering approach for improving the DHA accumulation in oleaginous microbes.

Species of Schizochytrium are well known for their remarkable ability to produce lipids intracellularly. However, during their lipid accumulation, reactive oxygen species (ROS) are generated inevitably as byproducts, which if in excess results in lipid peroxidation. To alleviate such ROS-induced damage, seven different natural antioxidants (ascorbic acid, α-tocopherol, tea extract, melatonin, mannitol, sesamol, and butylated hydroxytoluene) were evaluated for their effects on the lipid accumulation in Schizochytrium sp. PKU#Mn4 using a fractional factorial design. Among the tested antioxidants, mannitol showed the best increment (44.98%) in total fatty acids concentration. However, the interaction effects of mannitol (1 g/L) and ascorbic acid (1 g/L) resulted in 2.26 ± 0.27 g/L and 1.45 ± 0.04 g/L of saturated and polyunsaturated fatty acids (SFA and PUFA), respectively, in batch fermentation. These concentrations were further increased to 7.68 ± 0.37 g/L (SFA) and 5.86 ± 0.03 g/L (PUFA) through fed-batch fermentation. Notably, the interaction effects yielded 103.7% and 49.6% increment in SFA and PUFA concentrations in batch fermentation. The possible mechanisms underlining those increments were an increased maximum growth rate of strain PKU#Mn4, alleviated ROS level, and the differential expression of lipid biosynthetic genes andupregulated catalase gene. This study provides an applicable strategy for improving the accumulation of SFA and PUFA in thraustochytrids by exogenous antioxidants and the underlying mechanisms.

Thraustochytrids are well-known unicellular heterotrophic marine protists because of their promising ability to accumulate docosahexaenoic acid (DHA). However, the implications of their unique genomic and metabolic features on DHA production remain poorly understood. Here, the effects of chemical and physical culture conditions on the cell mass and DHA production were investigated for a unique thraustochytrid strain, PKU#SW8, isolated from the seawater of Pearl River Estuary. All the tested fermentation parameters showed a significant influence on the cell mass and concentration and yield of DHA. The addition of monosaccharides (fructose, mannose, glucose, or galactose) or glycerol to the culture medium yielded much higher cell mass and DHA concentrations than that of disaccharides and starch. Similarly, organic nitrogen sources (peptone, yeast extract, tryptone, and sodium glutamate) proved to be beneficial in achieving a higher cell mass and DHA concentration. PKU#SW8 was found to grow and accumulate a considerable amount of DHA over wide ranges of KH₂PO₄ (0.125–1.0 g/L), salinity (0–140% seawater), pH (3–9), temperature (16–36 °C), and agitation (140–230 rpm). With the optimal culture conditions (glycerol, 20 g/L; peptone, 2.5 g/L; 80% seawater; pH 4.0; 28 °C; and 200 rpm) determined based on the shake-flask experiments, the cell mass and concentration and yield of DHA were improved up to 7.5 ± 0.05 g/L, 2.14 ± 0.03 g/L, and 282.9 ± 3.0 mg/g, respectively, on a 5-L scale fermentation. This study provides valuable information about the fermentation conditions of the PKU#SW8 strain and its unique physiological features, which could be beneficial for strain development and large-scale DHA production.

Heterophily emerges as a critical challenge in Graph Anomaly Detection (GAD). Recent studies reveal that neighborhood distributions, rather than heterophily itself, are the fundamental factor for the expressive power of Graph Neural Networks (GNNs). However, two key challenges remain unresolved. First, the overlap in neighborhood distributions between anomalous and normal nodes poses significant difficulties in distinguishing them effectively. Second, the dispersion in neighborhood distributions within the same class prevents the application of a fixed aggregation strategy to accommodate the diverse patterns within the class. To tackle the aforementioned challenges, we propose a novel Graph Neural Network model called Neighborhood Adaptive Aggregation and Spectral Tuning (NAAST-GNN). Specifically, we first design a neighborhood adaptive aggregation module that adjusts the message passing mechanism based on the predicted probabilities for different node classes, ensuring that nodes from distinct classes but with similar neighborhood distributions derive unique aggregated neighborhood information. We then present a spectral tuning module that dynamically selects and combines spectral filters based on the predicted neighborhood distribution, ensuring adaptability to the diverse neighborhood distributions of nodes within the same class. Comprehensive experimental results demonstrate that our method outperforms state-of-the-art baselines.

In the task of code translation, neural network-based models have been shown to frequently produce semantically erroneous code that deviates from the original logic of the source code. This issue persists even with advanced large models. Although a recent approach proposed using test cases to identify these semantic errors, it relies heavily on the quality of the test cases and is not applicable to code snippets without test cases in real-world scenarios. Therefore, We present EISP, a static analysis framework based on the Large Language Model (LLM).First, the framework generates a semantic mapping between source code and translated code. Next, each sub-code fragment is identified by recursively traversing the abstract syntax tree of the source code, and its corresponding translated code fragment is found through the semantic mapping. Finally, EISP connects each pair of sub-code fragments with fine-grained knowledge hints through an AI chain to assist LLMs in discovering semantic mistakes in the translated code. In our benchmark evaluation, the EISP framework, based on GPT-4o mini, achieved an accuracy of 82.3\%, representing a 20.3\% improvement over baseline methods using the same base model, and a 7.4\% improvement compared to dynamic analysis methods that require test cases and manual intervention. To our knowledge, EISP is the first tool to locate semantic errors in translated code without test cases or compilable code. This innovative tool provides the software engineering community with a new way to deal with code fragments without test cases.

Code quality evaluation involves scoring generated code quality based on a reference code for a specific problem statement. Currently, there are two main forms of evaluating code quality: match-based evaluation and execution-based evaluation. The former requires the collection of a large number of test cases, making a huge cost. The latter relies on superficial code matching as an evaluation metric, which fails to accurately capture code semantics. Moreover, extensive research has demonstrated that match-based evaluations do not truly reflect code quality. With the development of large language models (LLMs) in recent years, studies have proven the feasibility of using LLMs as evaluators for generative tasks. However, due to issues like hallucinations and uncertainty in LLMs, their correlation with human judgment remains at a lower level, making the direct use of LLMs for code quality evaluation challenging. To address these issues, we propose Human-Like Code Quality Evaluation through LLM-based Recursive Semantic Comprehension (HuCoSC). We employ a recursive approach to enable LLMs to comprehend portions of code semantics independently each time, obtaining the code semantics through multiple interactions with LLMs. We designed a Semantic Dependency Decoupling Storage to make independent analysis feasible, allowing LLMs to achieve more accurate semantics by breaking down complex problems. Finally, the generated code is scored based on a semantic comparison between the reference code and itself. Experimental results indicate that HuCoSC surpasses existing state-of-the-art methods in terms of correlation with human experts and correlation with code execution.