Search for: All records

Abstract Heterologous expression of polyketide synthase (PKS) genes inEscherichia colihas enabled the production of various valuable natural and synthetic products. However, the limited availability of malonyl-CoA (M-CoA) inE. coliremains a substantial impediment to high-titer polyketide production. Here we address this limitation by disrupting the native M-CoA biosynthetic pathway and introducing an orthogonal pathway comprising a malonate transporter and M-CoA ligase, enabling efficient M-CoA biosynthesis under malonate supplementation. This approach substantially increases M-CoA levels, enhancing fatty acid and polyketide titers while reducing the promiscuous activity of PKSs toward undesired acyl-CoA substrates. Subsequent adaptive laboratory evolution of these strains provides insights into M-CoA regulation and identifies mutations that further boost M-CoA and polyketide production. This strategy improvesE. colias a host for polyketide biosynthesis and advances understanding of M-CoA metabolism in microbial systems. 

Fine-tuning pre-trained language models is a common practice in building NLP models for various tasks, including the case with less supervision. We argue that under the few-shot setting, formulating fine-tuning closer to the pre-training objective shall be able to unleash more benefits from the pre-trained language models. In this work, we take few-shot named entity recognition (NER) for a pilot study, where existing fine-tuning strategies are much different from pre-training. We propose a novel few-shot fine-tuning framework for NER, FFF-NER. Specifically, we introduce three new types of tokens, “is-entity”, “which-type” and “bracket”, so we can formulate the NER fine-tuning as (masked) token prediction or generation, depending on the choice of the pre-training objective. In our experiments, we apply to fine-tune both BERT and BART for few-shot NER on several benchmark datasets and observe significant improvements over existing fine-tuning strategies, including sequence labeling, prototype meta-learning, and prompt-based approaches. We further perform a series of ablation studies, showing few-shot NER performance is strongly correlated with the similarity between fine-tuning and pre-training.