An Improved U-Net Infrared Small Target Detection Algorithm Based on Multi-Scale Feature Decomposition and Fusion and Attention Mechanism.

Infrared small target detection technology plays a crucial role in various fields such as military reconnaissance, power patrol, medical diagnosis, and security. The advancement of deep learning has led to the success of convolutional neural networks in target segmentation. However, due to challenges like small target scales, weak signals, and strong background interference in infrared images, convolutional neural networks often face issues like leakage and misdetection in small target segmentation tasks. To address this, an enhanced U-Net method called MST-UNet is proposed, the method combines multi-scale feature decomposition and fusion and attention mechanisms. The method involves using Haar wavelet transform instead of maximum pooling for downsampling in the encoder to minimize feature loss and enhance feature utilization. Additionally, a multi-scale residual unit is introduced to extract contextual information at different scales, improving sensory field and feature expression. The inclusion of a triple attention mechanism in the encoder structure further enhances multidimensional information utilization and feature recovery by the decoder. Experimental analysis on the NUDT-SIRST dataset demonstrates that the proposed method significantly improves target contour accuracy and segmentation precision, achieving IoU and nIoU values of 80.09% and 80.19%, respectively.

The Formation of D-Allulose 3-Epimerase Hybrid Nanoflowers and Co-Immobilization on Resins for Improved Enzyme Activity, Stability, and Processability.

As a low-calorie sugar, D-allulose is produced from D-fructose catalyzed by D-allulose 3-epimerase (DAE). Here, to improve the catalytic activity, stability, and processability of DAE, we reported a novel method by forming organic-inorganic hybrid nanoflowers (NF-DAEs) and co-immobilizing them on resins to form composites (Re-NF-DAEs). NF-DAEs were prepared by combining DAE with metal ions (Co2+, Cu2+, Zn2+, Ca2+, Ni2+, Fe2+, and Fe3+) in PBS buffer, and were analyzed by scanning electron microscopy (SEM), Fourier transform infrared spectroscopy, and X-ray diffraction. All of the NF-DAEs showed higher catalytic activities than free DAE, and the NF-DAE with Ni2+ (NF-DAE-Ni) reached the highest relative activity of 218%. The NF-DAEs improved the thermal stability of DAE, and the longest half-life reached 228 min for NF-DAE-Co compared with 105 min for the free DAE at 55 °C. To further improve the recycling performance of the NF-DAEs in practical applications, we combined resins and NF-DAEs to form Re-NF-DAEs. Resins and NF-DAEs co-effected the performance of the composites, and ReA (LXTE-606 neutral hydrophobic epoxy-based polypropylene macroreticular resins)-based composites (ReA-NF-DAEs) exhibited outstanding relative activities, thermal stabilities, storage stabilities, and processabilities. The ReA-NF-DAEs were able to be reused to catalyze the conversion from D-fructose to D-allulose, and kept more than 60% of their activities after eight cycles.

The dual active sites reconstruction on gelatin in-situ derived 3D porous N-doped carbon for efficient and stable overall water splitting.

The exploration of bifunctional electrocatalysts with high activity, stability, and economy is of great significance in promoting the development of water splitting. Herein, a dual active sites heterostructure NiCoS/NC was designed to be derived in situ on 3D N-doped porous carbon (NC) using gelatin as a nitrogen and carbon source. The characterization of experiments suggests that nanoflower-like Ni2CoS4 (abbreviated as NiCoS) was randomly distributed on the NC substrate, and the sheet-like NC formed a highly open porous network structure resembling a honeycomb, which provided more accessible active sites for electrolyte ions. In addition, the special nanostructures of the catalyst materials help to promote the surface reconstruction to the real active substance NiOOH/CoOOH, and the double active sites synergistically reduce the overpotential of OER and improve its kinetics. DFT (Density-functional theory) calculations reveal the electronic coupling of NiCoS/NC in atomic orbitals, modulation of electrons by the heterointerface and N-doping, and synergistic effect of dual active sites improving the inherent catalytic activity. The NiCoS/NC composite electrocatalyst exhibited a 177 mV small OER overpotential and a 132 mV small HER overpotential with Faraday efficiencies as high as 96 % and 98 % at 10 mA cm-2 current density. In the two-electrode system, it also requires only an ultra-low voltage of 1.52 V to achieve a 10 mA cm-2 current density, and it shows excellent long-term water splitting stability. This provides a new idea for the development of transition metal-based bifunctional electrocatalysts.

Metabolic engineering of Saccharomyces cerevisiae for de novo production of odd-numbered medium-chain fatty acids.

Odd-numbered fatty acids (FAs) have been widely used in nutrition, agriculture, and chemical industries. Recently, some studies showed that they could be produced from bacteria or yeast, but the products are almost exclusively odd-numbered long-chain FAs. Here we report the design and construction of two biosynthetic pathways in Saccharomyces cerevisiae for de novo production of odd-numbered medium-chain fatty acids (OMFAs) via ricinoleic acid and 10-hydroxystearic acid, respectively. The production of OMFAs was enabled by introducing a hydroxy fatty acid cleavage pathway, including an alcohol dehydrogenase from Micrococcus luteus, a Baeyer-Villiger monooxygenase from Pseudomonas putida, and a lipase from Pseudomonas fluorescens. These OMFA biosynthetic pathways were optimized by eliminating the rate-limiting step, generating heptanoic acid, 11-hydroxyundec-9-enoic acid, nonanoic acid, and 9-hydroxynonanoic acid at 7.83 mg/L, 9.68 mg/L, 9.43 mg/L and 13.48 mg/L, respectively. This work demonstrates the biological production of OMFAs in a sustainable manner in S. cerevisiae.

Disruption of UDP-galactopyranose mutase expression: A novel strategy for regulation of galactomannan biosynthesis and monascus pigments secretion in Monascus purpureus M9.

The impact of the cell wall structure of Monascus purpureus M9 on the secretion of extracellular monascus pigments (exMPs) was investigated. To modify the cell wall structure, UDP-galactopyranose mutase (GlfA) was knocked out using Agrobacterium-mediated transformation method, leading to a significant reduction in the Galf-based polysaccharide within the cell wall. Changes in mycelium morphology, sporogenesis, and the expression of relevant genes in M9 were also observed following the mutation. Regarding MPs secretion, a notable increase was observed in six types of exMPs (R1, R2, Y1, Y2, O1 and O2). Specifically, these exMPs exhibited enhancement of 1.33, 1.59, 0.8, 2.45, 2.89 and 4.03 times, respectively, compared to the wild-type strain. These findings suggest that the alteration of the cell wall structure could selectively influence the secretion of MPs in M9. The underlying mechanisms were also discussed. This research contributes new insights into the regulation of the synthesis and secretion of MPs in Monascus spp..

Fusing Self-Attention and CoordConv to Improve the YOLOv5s Algorithm for Infrared Weak Target Detection.

Convolutional neural networks have achieved good results in target detection in many application scenarios, but convolutional neural networks still face great challenges when facing scenarios with small target sizes and complex background environments. To solve the problem of low accuracy of infrared weak target detection in complex scenes, and considering the real-time requirements of the detection task, we choose the YOLOv5s target detection algorithm for improvement. We add the Bottleneck Transformer structure and CoordConv to the network to optimize the model parameters and improve the performance of the detection network. Meanwhile, a two-dimensional Gaussian distribution is used to describe the importance of pixel points in the target frame, and the normalized Guassian Wasserstein distance (NWD) is used to measure the similarity between the prediction frame and the true frame to characterize the loss function of weak targets, which will help highlight the targets with flat positional deviation transformation and improve the detection accuracy. Finally, through experimental verification, compared with other mainstream detection algorithms, the improved algorithm in this paper significantly improves the target detection accuracy, with the mAP reaching 96.7 percent, which is 2.2 percentage points higher compared with Yolov5s.

Centromere Plasticity With Evolutionary Conservation and Divergence Uncovered by Wheat 10+ Genomes.

Centromeres (CEN) are the chromosomal regions that play a crucial role in maintaining genomic stability. The underlying highly repetitive DNA sequences can evolve quickly in most eukaryotes, and promote karyotype evolution. Despite their variability, it is not fully understood how these widely variable sequences ensure the homeostasis of centromere function. In this study, we investigated the genetics and epigenetics of CEN in a population of wheat lines from global breeding programs. We captured a high degree of sequences, positioning, and epigenetic variations in the large and complex wheat CEN. We found that most CENH3-associated repeats are Cereba element of retrotransposons and exhibit phylogenetic homogenization across different wheat lines, but the less-associated repeat sequences diverge on their own way in each wheat line, implying specific mechanisms for selecting certain repeat types as functional core CEN. Furthermore, we observed that CENH3 nucleosome structures display looser wrapping of DNA termini on complex centromeric repeats, including the repositioned CEN. We also found that strict CENH3 nucleosome positioning and intrinsic DNA features play a role in determining centromere identity among different lines. Specific non-B form DNAs were substantially associated with CENH3 nucleosomes for the repositioned centromeres. These findings suggest that multiple mechanisms were involved in the adaptation of CENH3 nucleosomes that can stabilize CEN. Ultimately, we proposed a remarkable epigenetic plasticity of centromere chromatin within the diverse genomic context, and the high robustness is crucial for maintaining centromere function and genome stability in wheat 10+ lines as a result of past breeding selections.

De novo bio-production of odd-chain fatty acids in Saccharomyces cerevisiae through a synthetic pathway via 3-hydroxypropionic acid.

Odd-chain fatty acids (OCFAs) and their derivatives have attracted increasing attention due to their wide applications in the chemical, fuel, and pharmaceutical industry. However, most natural fatty acids are even-chained, and OCFAs are rare. In this work, a novel pathway was designed and established for de novo synthesis of OCFAs via 3-hydroxypropionic acid (3-HP) as the intermediate in Saccharomyces cerevisiae. First, the OCFAs biosynthesis pathway from 3-HP was confirmed, followed by an optimization of the precursor 3-HP. After combining these strategies, a de novo production of OCFAs at 74.8 mg/L was achieved, and the percentage of OCFAs in total lipids reached 20.3%, reaching the highest ratio of de novo-produced OCFAs. Of the OCFAs produced by the engineered strain, heptadecenoic acid (C17:1) and heptadecanoic acid (C17:0) accounted for 12.1% and 7.6% in total lipid content, respectively. This work provides a new and promising pathway for the de novo bio-production of OCFAs.

Decentralized Primal-Dual Proximal Operator Algorithm for Constrained Nonsmooth Composite Optimization Problems over Networks.

In this paper, we focus on the nonsmooth composite optimization problems over networks, which consist of a smooth term and a nonsmooth term. Both equality constraints and box constraints for the decision variables are also considered. Based on the multi-agent networks, the objective problems are split into a series of agents on which the problems can be solved in a decentralized manner. By establishing the Lagrange function of the problems, the first-order optimal condition is obtained in the primal-dual domain. Then, we propose a decentralized algorithm with the proximal operators. The proposed algorithm has uncoordinated stepsizes with respect to agents or edges, where no global parameters are involved. By constructing the compact form of the algorithm with operators, we complete the convergence analysis with the fixed-point theory. With the constrained quadratic programming problem, simulations verify the effectiveness of the proposed algorithm.

Lysophosphatidylcholine acyltransferase 1 promotes epithelial-mesenchymal transition of hepatocellular carcinoma via the Wnt/ß-catenin signaling pathway.

INTRODUCTION AND OBJECTIVES: Hepatocellular carcinoma (HCC) is one of the most malignant digestive tumors, and its insidious onset and rapid progression are the main reasons for the difficulty in effective treatment. Lysophosphatidylcholine acyltransferase 1 (LPCAT1) is a key enzyme that regulates phospholipid metabolism of the cell membrane. However, the mechanism by which LPCAT1 regulates HCC metastasis remains unknown. This study aimed to explore its biological function and potential mechanisms concerning migration and invasion in HCC. MATERIALS AND METHODS: LPCAT1 expression in HCC tissues and its association with clinical outcomes were investigated by western blotting and bioinformatic methods, respectively. The role of LPCAT1 in migration and invasion was assessed via Transwell assays. The expression pattern of epithelial-mesenchymal transition (EMT) markers was quantified by western blotting. The biological behaviors of LPCAT1 in vivo were evaluated using xenograft tumor models and caudal vein metastatic models. Signaling pathways related to LPCAT1 were predicted using gene set enrichment analysis (GSEA) and further confirmed by western blotting. RESULTS: LPCAT1 expression was significantly upregulated in HCC tissues and indicated a poor prognosis of HCC patients. Several EMT-related markers were found to be regulated by LPCAT1. HCC cells overexpressing LPCAT1 exhibited remarkably high migration and invasion capacities, upregulated expression of mesenchymal markers and reduced E-cadherin expression. In vivo, LPCAT1 promoted HCC pulmonary metastasis. Furthermore, the Wnt/ß-catenin signaling pathway was confirmed to be activated by LPCAT1. CONCLUSIONS: LPCAT1 could serve as a promising biomarker of HCC and as a novel therapeutic target for the treatment of metastatic HCC.

Metabolic engineering of threonine catabolism enables Saccharomyces cerevisiae to produce propionate under aerobic conditions.

BACKGROUND: Propionate is widely used as a preservative in the food and animal feed industries. Propionate is currently produced by petrochemical processes, and fermentative production of propionate remains challenging. METHODS AND RESULTS: In this study, a synthetic propionate pathway was constructed in the budding yeast Saccharomyces cerevisiae, for propionate production under aerobic conditions. Through expression of tdcB and aldH from Escherichia coli and kivD from Lactococcus lactis, L-threonine was converted to propionate via 2-ketobutyrate and propionaldehyde. The resulting yeast aerobically produced 0.21 g L-1 propionate from glucose in a shake flask. Subsequent overexpression of pathway genes and elimination of competing pathways increased propionate production to 0.37 g L-1 . To further increase propionate production, carbon flux was pulled into the propionate pathway by weakened expression of pyruvate kinase (PYK1), together with overexpression of phosphoenolpyruvate carboxylase (ppc). The final propionate production reached 1.05 g L-1 during fed-batch fermentation in a fermenter. CONCLUSIONS AND IMPLICATIONS: In this work, a yeast cell factory was constructed using synthetic biology and metabolic engineering strategies to enable propionate production under aerobic conditions. Our study demonstrates engineered S. cerevisiae as a promising alternative for the production of propionate and its derivatives.

Genome-wide association studies of plant architecture-related traits and 100-seed weight in soybean landraces.

BACKGROUND: Plant architecture-related traits (e.g., plant height (PH), number of nodes on main stem (NN), branch number (BN) and stem diameter (DI)) and 100-seed weight (100-SW) are important agronomic traits and are closely related to soybean yield. However, the genetic basis and breeding potential of these important agronomic traits remain largely ambiguous in soybean (Glycine max (L.) Merr.). RESULTS: In this study, we collected 133 soybean landraces from China, phenotyped them in two years at two locations for the above five traits and conducted a genome-wide association study (GWAS) using 82,187 single nucleotide polymorphisms (SNPs). As a result, we found that a total of 59 SNPs were repeatedly detected in at least two environments. There were 12, 12, 4, 4 and 27 SNPs associated with PH, NN, BN, DI and 100-SW, respectively. Among these markers, seven SNPs (AX-90380587, AX-90406013, AX-90387160, AX-90317160, AX-90449770, AX-90460927 and AX-90520043) were large-effect markers for PH, NN, BN, DI and 100-SW, and 15 potential candidate genes were predicted to be in linkage disequilibrium (LD) decay distance or LD block. In addition, real-time quantitative PCR (qRT-PCR) analysis was performed on four 100-SW potential candidate genes, three of them showed significantly different expression levels between the extreme materials at the seed development stage. Therefore, Glyma.05 g127900, Glyma.05 g128000 and Glyma.05 g129000 were considered as candidate genes with 100-SW in soybean. CONCLUSIONS: These findings shed light on the genetic basis of plant architecture-related traits and 100-SW in soybean, and candidate genes could be used for further positional cloning.

Development of Host-Orthogonal Genetic Systems for Synthetic Biology.

The construction of a host-orthogonal genetic system can not only minimize the impact of host-specific nuances on fine-tuning of gene expression, but also expand cellular functions such as in vivo continuous evolution of genes based on an error-prone DNA polymerase. It represents an emerging powerful approach for making biology easier to engineer. In this review, the recent advances are described on the design of genetic systems that can be stably inherited in the host cells and are responsible for important biological processes including DNA replication, RNA transcription, protein translation, and gene regulation. Their applications in synthetic biology are summarized and the future challenges and opportunities are discussed in developing such systems.

Effect of Qizhitongluo capsule on lower limb rehabilitation after stroke: A randomized clinical trial.

BACKGROUND: An individual's level of lower limb motor function is associated with his or her disability level after stroke, and motor improvement may lead to a better prognosis and quality of life. Data from animal models show that Qizhitongluo (QZTL) capsule facilitates recovery after focal brain injury. We aimed to validate the efficacy and safety of the QZTL capsule for promoting lower limb motor recovery in poststroke patients. METHODS: In this randomized, multicenter, double-blind, placebo- and active-controlled trial from 13 sites in China, participants with ischemic stroke and Fugl-Meyer motor scale (FMMS) scores of <95 were eligible for inclusion. Patients were randomly assigned in a 2:1:1 ratio to the QZTL group, Naoxintong (NXT) group or placebo group for 12 weeks at 15-28 days after the onset of stroke. The primary outcome was the change in the Lower Limb FMMS (FMMS-LL) score from baseline over the 12-week intervention period. RESULTS: 622 participants were randomly assigned to the QZTL group (309), NXT group (159), or placebo group (154). The FMMS-LL score increased by 4.81 points (95 % CI, 4.27-5.35) in the QZTL group, by 3.77 points (95 % CI, 3.03-4.51) in the NXT group and by 3.00 points (95 % CI, 3.03-4.51) in the placebo group at week 12. The QZTL group showed significantly larger improvements compared with the placebo group at each interview from weeks 4-12 (difference, 0.89 [0.30,1.49] at week 4, P = 0.0032; difference, 1.83[1.01,2.66] at 90 days poststroke, P < 0.0001; difference, 1.81[0.88,2.74] at week 12, P = 0.0001). CONCLUSION: The QZTL capsule is an effective treatment for lower limb motor impairment. The finding indicates that the QZTL capsule may be used as a potential new strategy for stroke rehabilitation.

Development and Validation of a Novel Hypoxia-Related Long Noncoding RNA Model With Regard to Prognosis and Immune Features in Breast Cancer.

Background: Female breast cancer is currently the most frequently diagnosed cancer in the world. This study aimed to develop and validate a novel hypoxia-related long noncoding RNA (HRL) prognostic model for predicting the overall survival (OS) of patients with breast cancer. Methods: The gene expression profiles were downloaded from The Cancer Genome Atlas (TCGA) database. A total of 200 hypoxia-related mRNAs were obtained from the Molecular Signatures Database. The co-expression analysis between differentially expressed hypoxia-related mRNAs and lncRNAs based on Spearman's rank correlation was performed to screen out 166 HRLs. Based on univariate Cox regression and least absolute shrinkage and selection operator Cox regression analysis in the training set, we filtered out 12 optimal prognostic hypoxia-related lncRNAs (PHRLs) to develop a prognostic model. Kaplan-Meier survival analysis, receiver operating characteristic curves, area under the curve, and univariate and multivariate Cox regression analyses were used to test the predictive ability of the risk model in the training, testing, and total sets. Results: A 12-HRL prognostic model was developed to predict the survival outcome of patients with breast cancer. Patients in the high-risk group had significantly shorter median OS, DFS (disease-free survival), and predicted lower chemosensitivity (paclitaxel, docetaxel) compared with those in the low-risk group. Also, the risk score based on the expression of the 12 HRLs acted as an independent prognostic factor. The immune cell infiltration analysis revealed that the immune scores of patients in the high-risk group were lower than those of the patients in the low-risk group. RT-qPCR assays were conducted to verify the expression of the 12 PHRLs in breast cancer tissues and cell lines. Conclusion: Our study uncovered dozens of potential prognostic biomarkers and therapeutic targets related to the hypoxia signaling pathway in breast cancer.

Development and Application of CRISPR/Cas in Microbial Biotechnology.

The clustered regularly interspaced short palindromic repeats (CRISPR)-associated (Cas) system has been rapidly developed as versatile genomic engineering tools with high efficiency, accuracy and flexibility, and has revolutionized traditional methods for applications in microbial biotechnology. Here, key points of building reliable CRISPR/Cas system for genome engineering are discussed, including the Cas protein, the guide RNA and the donor DNA. Following an overview of various CRISPR/Cas tools for genome engineering, including gene activation, gene interference, orthogonal CRISPR systems and precise single base editing, we highlighted the application of CRISPR/Cas toolbox for multiplexed engineering and high throughput screening. We then summarize recent applications of CRISPR/Cas systems in metabolic engineering toward production of chemicals and natural compounds, and end with perspectives of future advancements.

The effectiveness and safety of acupuncture for the treatment of myasthenia gravis: a systematic review and meta-analysis of randomized controlled trials.

BACKGROUND: No systematic reviews of acupuncture as a treatment for myasthenia gravis (MG) have been published in English. The aim of our study is to evaluate the efficacy and safety of acupuncture as a treatment for MG. METHODS: We searched for randomized controlled trials (RCTs) in seven main electronic databases. Unpublished articles, including conference papers and Chinese doctoral and master's theses, were also included as supplementary sources. The primary outcome was the relative clinical score (RCS) response rate. We performed a meta-analysis using RR and MD with 95% CI. RESULTS: Thirteen RCTs involving a total of 775 participants were included. Most included trials had a high risk of bias in allocation concealment and blinding. Eleven RCTs used acupuncture as an adjuvant to medication, and this treatment showed a significant improvement in the RCS response rate compared to medication alone (RR: 1.42; 95% CI: 1.06-1.91; P=0.02). The subgroup analysis based on the treatment duration showed a significant effect on the RCS response rate when the treatment duration was longer than 12 weeks (RR: 2.02; 95% CI: 1.31-3.12; P=0.001). In contrast, there was no significant effect of treatment with a duration less than 8 weeks (RR: 1.14; 95% CI: 0.91-1.44; P=0.26). Four RCTs showed a significant difference in the absolute clinical score (ACS) (RR: 3.42; 95% CI: 1.23-5.61; P=0.002). The acupuncture group reported better outcomes. No severe adverse events corresponding to acupuncture were reported. CONCLUSIONS: This meta-analysis suggests that acupuncture as an integrative therapy has a significant positive effect in treating MG. Acupuncture may enhance the efficacy of medication in MG patients. The safety of acupuncture requires further investigation. The clinical significance of these changes needs to be investigated by further studies using rigorous designs and longer follow-up times.

Treatment of migraines with Tianshu capsule: a multi-center, double-blind, randomized, placebo-controlled clinical trial.

BACKGROUND: Tianshu capsule (TSC), a formula of traditional Chinese medicine, has been widely used in clinical practice for prophylactic treatment of headaches in China. However, former clinical trials of TSC were small, and lack of a standard set of diagnostic criteria to enroll patients. The study was conducted to re-evaluate the efficacy and safety of TSC post-marketing in an extending number of migraineurs who have diagnosed migraine with the International Classification of Headache Disorders, 3rd edition (beta version, ICHD-3ß). METHODS: The study was a double-blind, randomized, placebo-controlled clinical trial that conducted at 20 clinical centers in China. At enrollment, patients between 18 and 65 years of age diagnosed with migraine were assigned to receive either TSC (4.08 g, three times daily) or a matched placebo according to a randomization protocol. The primary endpoint was a relative reduction of 50% or more in the frequency of headache attacks. The secondary outcomes included a reduction in the incidence of headache, the visual analogue scale of headache attacks, days of acute analgesic usage, and percentage of patients with a decrease of 50% or more in headache severity. Accompanying symptoms were also assessed. RESULTS: One thousand migraine patients were initially enrolled in the study, and 919 of them completed the trial. Following the 12-week treatment, significant improvement was observed in the TSC group concerning both primary and secondary outcomes. After therapy discontinuation, the gap between the TSC group and the placebo group in efficacy outcomes continued to increase. There were no severe adverse effects. CONCLUSIONS: TSC is an effective, well-tolerated medicine for prophylactic treatment of migraine, and still have prophylactic effect after medicine discontinuation. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT02035111; Data of registration: 2014-01-10.

Engineering the 5' UTR-Mediated Regulation of Protein Abundance in Yeast Using Nucleotide Sequence Activity Relationships.

The 5' untranslated region (5'UTR) plays a key role in post-transcriptional regulation, but interaction between nucleotides and directed evolution of 5'UTRs as synthetic regulatory elements remain unclear. By constructing a library of synthesized random 5'UTRs of 24 nucleotides in Saccharomyces cerevisiae, we observed strong epistatic interactions among bases from different positions in the 5'UTR. Taking into account these base interactions, we constructed a mathematical model to predict protein abundance with a precision of R2 = 0.60. On the basis of this model, we developed an approach to engineer 5'UTRs according to nucleotide sequence activity relationships (NuSAR), in which 5'UTRs were engineered stepwise through repeated cycles of backbone design, directed screening, and model reconstruction. After three rounds of NuSAR, the predictive accuracy of our model was improved to R2 = 0.71, and a strong 5'UTR was obtained with 5-fold higher protein abundance than the starting 5'UTR. Our findings provide new insights into the mechanism of 5'UTR regulation and  contribute to a new translational elements engineering approach in synthetic biology.