Dual-model deep learning for Alzheimer's prognostication.

PtRu nanoparticle catalysts with adjustable electronic environments for efficient low-temperature dehydrogenation of cycloalkanes.

The purpose of this study was to investigate the optimal cultivation strategy for long-term stable ethanol production from sweet potato starch in vertical mass flow bioreactor (VMFB) by co-immobilizing Aspergillus awamori, Rhizopus japonicus and Zymomonas mobilis. In exploring the optimum initial substrate concentration, the maximum ethanol concentration of 45.11 g/L was achieved at the 54th hour at a feed concentration of 12 % (w/v), which resulted in a conversion rate of 66.26 % compared to the theoretical value. In the repeated batch process (multiple cycles), stable ethanol production was achieved from 6 % (w/v, 3cycles) to 12 % (w/v, 5cycles) of sweet potato starch producing total ethanol concentrations of 71.03 g/L (3cycles, 72 h) and 241.54 g/L (5 cycles, 348 h), respectively. In fed-batch fermentation with a total addition of 24 % (w/v) sweet potato starch (four separate additions), the maximum ethanol concentration and yield after 114 h were 102.6 g/L and 0.43, respectively. To evaluate the potential for future industrialization, dried raw starch was used instead of gelatinized starch as a substrate. The results demonstrated that the fed-batch fermentation process had higher ethanol yield and starch conversion rates, with values of 0.32 and 55.96 % at 144 h, respectively.

Structural evolution-based ancestral sequence reconstruction of medium-chain alcohol dehydrogenase for efficient synthesis of fused-ring chiral alcohols.

To compare the short-term outcomes of robotic distal pancreatectomy (RDP) and laparoscopic distal pancreatectomy (LDP), with a particular focus on the actual completion rates of spleen and splenic vessels preservation preoperatively. Fifty-two RDP and 87 LDP procedures performed between 2014 and 2024 were retrospectively analyzed. The operative time was significantly longer for RDP than for LDP (median 407min vs. 320min, p < 0.0001), whereas the intraoperative blood loss was significantly lower (median 70 mL vs. 100 mL, p = 0.011). The rate of conversion to open surgery was also lower for RDP (2% vs. 10%, p = 0.042). The spleen preservation rate, defined as the proportion of cases in which spleen preservation was achieved preoperatively, was 100% (12/12) and 92% (23/25) in the RDP and LDP groups, respectively (p = 0.203). The splenic vessels preservation rate was significantly higher with RDP (100% [10/10]) than with LDP (67% [16/24]; p = 0.011). Postoperative complication rates did not differ significantly between the two groups. Compared with LDP, RDP was associated with significantly reduced intraoperative blood loss, a lower conversion rate to open surgery, and a higher success rate in splenic vessels preservation, suggesting the technical advantages of RDP.

Engineering Escherichia coli cell Factories for continuous 5'-cytidine monophosphate production via biofilm-anchored dual-enzyme cascade catalysis.

The aim of this study is to evaluate the influence of age on net energy and dietary phytase supplementation on the standardized ileal digestibility (SID) of amino acids of palm kernel meal (PKM). To determine the SID, A total of 162 12-day-old and 72 32-day-old male Arbor Acres chickens were divided into three groups (D0, D1 and D2). For the net energy a total of 72 11- and 31-day-old broiler chickens were divided into two groups (R1 and R2). Each group consisted of six repetitions and was fed its respective ration. The group D0 received a diet without nitrogen, then group D1 received an experimental diet with PKM as the sole nitrogen source, and finally group D2 received a diet without mineral phosphorus while supplemented with 1,000 FTU phytase/kg diet. For net energy trial, group R1 received a basal diet, while group R2 received a diet consisting of 70%R1 and 30% PKM. Bird performance, AME, NE, and N balance were measured in a calorimetry system. The results showed that there was no interaction (P > 0.05) between the age of the birds (20 versus 40 days) and the dietary treatments (D1 and D2) for the AID and SID of most Aas, except for the values of aspartate (Asp) regarding AID and SID. The trends of age effect on AA digestibility were used and the effect of diets on AA digestibility was similar to the AID of AA values. Finally, results regarding the effective energy value and energy conversion efficiency of PKM revealed that age significantly affected AME, NE, and AME/GE of PKM (P < 0.05). Additionally, the conversion rates of NE and AME of PKM were 72.22% and 79.98% during the starter and finisher phases, respectively. In conclusion, the results indicate that age affects the NE of PKM and the supplementation of phytase did not improve the digestibility of PKM amino acids.

Electrochemical purification of a cobalt-catecholate framework to suppress non-radiative recombination boosts photocatalytic CO2 reduction.

The increasing reliance on digital marketing has made it essential for organizations to evaluate the effectiveness of their campaigns and understand customer behavior. Traditional methods often depend on basic metrics such as clicks and impressions, which do not provide a complete picture of customer engagement or conversion outcomes. This project addresses this challenge by developing a machine learning-based system that predicts customer response to marketing campaigns and analyzes campaign performance. The system incorporates data preprocessing, feature engineering, and classification techniques to extract meaningful patterns from the data. By leveraging these techniques, the model provides probability-based predictions that help identify whether a customer is likely to respond to a campaign. In addition to prediction, the system includes an interactive dashboard that enables users to input campaign parameters and visualize key performance indicators such as conversion rate, return on investment (ROI), and trend analysis. Features like what-if simulation allow users to explore different marketing strategies and their potential impact. The integration of frontend, backend, and API components ensures smooth data flow and real-time interaction. Overall, the proposed system demonstrates how machine learning can be effectively applied to marketing analytics, supporting data-driven decision-making and improving the efficiency of marketing campaigns.

This paper proposes a model to estimate economic losses attributable to front-end complexity and web performance debt in web applications. Technical debt consequences arising from excessive JavaScript execution, deep DOM hierarchies, and high HTTP request volumes were examined. These factors degrade web performance, reduce user engagement and conversion rates, and ultimately diminish revenue. Using an analytical modelling approach, correlations were established between front-end complexity metrics, including JavaScript bundle size, DOM depth, and HTTP request volume, and web performance indicators, specifically Largest Contentful Paint (LCP) and Interaction to Next Paint (INP). Revenue and conversion losses were modeled as nonlinear functions of latency, incorporating margins, control intensity, and threshold effects to represent realistic performance-revenue relationships. The empirical evaluation relied on a three-layered dataset combining real-world performance metrics from Lighthouse with synthetic datasets modeling optimized (best case), typical (baseline), and high-complexity (worst case) front-end scenarios. The results confirmed that increased front-end complexity and performance debt correlate with deteriorated latency and interactivity, leading to substantial conversion and revenue losses. Marginal and threshold analyses revealed nonlinear effects: at lower complexity levels, performance improvements yield higher financial returns, whereas at higher complexity levels, optimisation produces diminishing marginal returns. These findings demonstrate that front-end performance optimisation is an economic imperative rather than a technical consideration. Effective management of front-end complexity reduces performance debt and revenue erosion, providing a framework for engineering decisions and investment strategies in performance-critical environments. This approach transforms performance management into a strategic economic decision, enabling investment optimisation through direct correlation with business outcomes.

Luteolin is a plant-derived flavone widely distributed in edible plants and medicinal herbs, attracting attention due to its antioxidant and anti-inflammatory properties. However, efficient biosynthesis of specific luteolin glycosides, particularly luteolin-5-O-xyloside, remains challenging due to low donor specificity, limited catalytic efficiency of natural glycosyltransferases, and the high cost of UDP-xylose (UDP-Xyl). In this study, a novel 5-O-xylosyltransferase and an engineered Escherichia coli whole-cell system were developed for luteolin-5-O-xyloside production. Structure-guided engineering identified Thr142 as a key residue for UDP-sugar recognition. The mutant ZmDo-T142 V showed enhanced catalytic efficiency toward UDP-Xyl (kcat/Km = 1302.8 s-1 M-1), 2.3-fold higher than the wild type. A de novo UDP-Xyl biosynthetic pathway was constructed, enabling intracellular cofactor regeneration. Under optimized conditions, a titer of 4451 mg/L with an 87% conversion rate was achieved.

Dihydrocarvone is an important flavor compound and chiral building block obtained through the ene reductase (ER)-mediated reduction of carvone. In this study, a new ER, KlebER1, from Klebsiella sp. O852 was identified and characterized. Optimal enzyme activity occurred at 40 °C and pH 7, with approximately 80% activity retained after incubation at 20-40 °C for 3 h. Based on sequence alignment, alanine scanning, and molecular docking analyses, a site-directed mutation (A303N) was introduced, resulting in a 1.73-fold enhancement in activity. Finally, an enzymatic cascade system incorporating the A303N mutant and glucose dehydrogenase for coenzyme regeneration was constructed both in vivo and in vitro for dihydrocarvone production. This biocatalytic system achieved a dihydrocarvone concentration of 5.88 mmol/L and a conversion rate of 99%, representing a significant improvement compared to the single use of A303N mutant. This study lays the foundation for investigating the function and modification of ERs and promoting the efficient biosynthesis of dihydrocarvone.

The Soybean aphid (Aphis glycines) is a destructive pest that threatens soybeans. In order to develop green and effective control strategies, we propose an EQPAL epidemic model that integrates four developmental stages (1st–2nd stage nymphs, 3rd stage nymphs, 4th stage nymphs, and adults) and a ladybug (Harmonia axyridis) compartment. This model achieves green pest control by artificially releasing a natural enemy of soybean aphids to prey on adult soybean aphids. We analyzed the dynamic behavior of the model and derived the basic reproduction number R0. Using field monitoring data from Changchun City, Jilin Province, China in 2025, the segmented nonlinear least squares method was used for parameter estimation and fitting, resulting in an overall determination coefficient of R2=0.8204. The numerical simulation results showed that the release of ladybugs significantly reduced the density and peak value of soybean aphid adults, and the predation rate β, predation conversion rate c, and ladybug migration rate ω were identified as key regulatory parameters. In addition, a cost–benefit analysis was conducted to determine the most cost-effective control measures.

This systematic review with meta-analysis aimed to compare the accuracy correction of lateral closing wedge (LCW) and medial opening wedge (MOW) high tibial osteotomies (HTO) in patients with knee osteoarthritis (OA), and to evaluate the proportion of complications, reinterventions, revisions and conversion to knee arthroplasty. PubMed, EMBASE and Web of Science were searched until April 2024 to identify studies reporting imaging measurements, complications, reinterventions, revisions, and conversion to knee arthroplasty after HTO for knee OA. Meta-analysis was performed to compare LCW and MOW. Meta-analyses are reported as mean differences (MD) for continuous outcomes, or proportions and risk ratio (RR) for dichotomous outcomes, with 95% confidence intervals (CI). A total of 103 studies (MOW, 43 studies; LCW, 33 studies; both techniques, 27 studies) were included comprising 12,200 knees from 11,472 patients (50.2% females, 52.4 ± 6.5 years old and 21.5 ± 12.2 kg/m²). Imaging measurements identified no significant differences between LCW and MOW in coronal plane alignment. Posterior tibial slope (PTS) was significantly higher in MOW as compared to the LCW (MD = 3.72°, 95% confidence interval [CI] 2.31°-5.13°). Proportion of complications was high at 22.2% (95% CI 17.6%-27.0%) and the conversion to knee arthroplasty was 6.4% (95% CI 2.7%-11.3%) at 5 years, 13.8% (95% CI 7.8%-21.1%) at 10 years and 33.4% (95% CI 15.0%-54.6%) at 15 years. There were no differences between MOW and LCW in risk of complications or proportion of conversion to knee arthroplasty, but risk of revision was lower in MOW (RR = 0.34, 95% CI 0.19-0.61). The MOW and LCW were comparable to correct the lower limb coronal plane alignment, but the LCW was superior in reducing the PTS angle. Both techniques showed a similar risk profile for complications, reintervention and conversion to knee arthroplasty, but the MOW had a 66% lower risk for revision. Level IV.

N-doping has been taken as a pivotal method for modulating the photocatalytic activity of covalent organic frameworks (COFs). To elucidate the relationship between the nitrogen position and the photocatalytic performance in one-dimensional (1D) COFs, three 1D COFs (Pd-COF, PdA-COF, and Pz-COF) (Pd-pyrimidine, PdA-asymmetric pyrimidine, Pz-pyrazine) were synthesized by different N-heterocycle-functionalized nonlinear C2 linkage (with different N positions) and planar C4 knots. Pd-COF and PdA-COF incorporate N atoms in the form of pyrimidine, while Pz-COF contains pyrazine units. Pd-COF and PdA-COF demonstrated superior photocatalytic hydrogen generation activity compared to Pz-COF, with rates of 99.6 and 44.6 mmol g-1 h-1, respectively, versus 18.0 mmol g-1 h-1 for Pz-COF. However, for the oxidation of hydroxyl groups from arylboronic acids, the Pz-COF and PdA-COF with N atoms introduced into the pores are more suitable. The conversion rates of Pz-COF (up to 99%) and PdA-COF (up to 90%) are significantly higher than that of PA-COF (up to 85%). This may result from the position of the N atom affecting the electron cloud density of the aromatic ring, which facilitates the different intramolecular charge transfers. Furthermore, N atoms located within the pore could also increase the active sites, promote the contact between arylboronic acid and active sites, thus improving the catalytic efficiency. This study provides an effective strategy for designing COFs-based photocatalysts with efficient photocatalytic hydrogen generation ability and emphasizes the potential of N-doping 1D COFs in the field of photocatalysis.

BackgroundMinimally invasive direct coronary artery bypass (MIDCAB) has emerged as an alternative to conventional coronary artery bypass grafting; however, its adoption remains limited due to technical complexity and a steep learning curve, particularly in patients requiring multi-vessel revascularization. Objective data defining the learning curve of multi-vessel MIDCAB are scarce.MethodsThis retrospective study included consecutive patients who underwent multi-vessel MIDCAB between January 2020 and December 2025. Patients requiring single-vessel revascularization were intentionally excluded to ensure procedural homogeneity. The learning curve was evaluated using cumulative sum (CUSUM) analysis, and cases were stratified into three phases based on CUSUM inflection points. Perioperative and postoperative outcomes were compared across learning curve phases.ResultsA total of 169 patients were analyzed. CUSUM analysis identified three distinct learning phases: an initial learning phase (cases 1-48), a transition phase (cases 49-107), and a proficiency phase (cases 108-169). With increasing surgical experience, cardiopulmonary bypass time, aortic cross-clamp time, and total operative duration decreased significantly. The rate of conversion to open surgery declined markedly across learning phases, whereas in-hospital mortality and major postoperative complications remained low and comparable. These findings indicate improved procedural efficiency without compromising early clinical outcomes.ConclusionsMulti-vessel MIDCAB is associated with a substantial learning curve that can be objectively characterized using CUSUM analysis. Surgical proficiency is achieved only after a considerable number of cases, emphasizing the importance of adequate case volume and structured performance monitoring. These results provide a practical benchmark for centers aiming to adopt or expand multi-vessel MIDCAB programs.

α-Galactosidase catalyzes the transglycosylation reaction for galactooligosaccharides (GOS) synthesis. However, its strong hydrolytic activity limits the yield of oligosaccharides. Here, semirational design was used to engineer α-galactosidase AglB from Lactobacillus amylolyticus L6, generating three mutants with improved transglycosylation. K480R and H203P exhibited different functional characteristics. The hydrolytic activity of the K480R mutant was significantly reduced by 90.37%, thereby increasing the maximum GOS conversion rate to 34.30%, marking a 3.32-fold increase compared to WT. In contrast, the hydrolytic activity of the H203P mutant was increased by 2.83-fold. Molecular dynamics simulations and substrate-binding pocket analysis revealed that the K480R mutation rendered the active pocket more rigid and shallower, which hindered the access of water molecules through steric hindrance. Conversely, the H203P mutation enhanced the local flexibility at the pocket entrance, thus facilitating nonspecific hydrolysis. This study shifted the functional balance of AglB from hydrolysis toward transglycosylation, providing a high-performance catalyst for GOS production.

ABSTRACT Guideline-based combination therapy achieves sputum culture conversion rates in 23%–34% of patients with Mycobacterium abscessus -complex lung disease. Thus, new therapies are needed. We performed a systematic review to validate and benchmark the hollow fiber system model of M. abscessus lung disease for drug development. We performed a literature search to identify all published hollow fiber system pharmacokinetic-pharmacodynamic studies. Preferred Reporting Items for Systematic Reviews and Meta-Analyses was used for bias minimization. A total of 12 studies were identified. The average quality score was 13.7 out of 21. Eight were monotherapy (exposure-effect and dose-fractionation), one double β-lactam, and three guideline-based therapy studies. For omadacycline and imipenem, hollow fiber system data were accompanied by clinical real-world evidence confirmation. Microbial kill was always terminated by antimicrobial resistance. We used quantitative analyses to rank drugs’ efficacy based on CFU/mL kill below day 0 bacterial burden normalized to multi-drug guideline-based therapy kill. The highest-ranked drugs were sulbactam-durlobactam-ceftriaxone (177-fold), epetraborole (15-fold), and omadacycline (7-fold) better than guideline-based therapy. We used the target exposures identified in the systematic analysis in Monte Carlo experiments to identify optimal doses for inhaled formulations. The optimal inhalational dose of imipenem was 250 mg/day, for tigecycline 4 mg/day, for cefoxitin 50 mg/day, and for amikacin liposome inhalation suspension 590 mg once weekly. The hollow fiber system model of M. abscessus lung disease is tractable for exposure-effect, dose-fractionation, and factorial design combination studies. It could also be used to rank drugs and inform on which drugs to test in novel combinations. IMPORTANCE Current treatments for Mycobacterium abscessus lung disease fail in 70%–80% of patients and are toxic. The hollow fiber system has been used to study old and new potential treatments for this disease. We performed a systematic review of this methodology, for lessons learned. We found 12 studies, which were of adequate quality. Efficacy was always terminated by antimicrobial resistance. The top three drugs in terms of efficacy were sulbactam-durlobactam-ceftriaxone, epetraborole, and omadacycline, which were 7 to 177 times better than standard of care. These drugs could be combined into a new treatment regimen better than current treatments. We also calculated new doses for imipenem, tigecycline, cefoxitin, and amikacin when administered as inhalational therapy. The inhaled doses were multiple-fold lower than intravenous ones, which could be less toxic. The hollow fiber system model is an easily managed system from drug development and dose finding for M. abscessus lung disease.

The electrochemical transformation of carbon dioxide into valuable fuels represents a critical pathway for sustainable energy cycles. A principal obstacle in this field is the creation of affordable and effective electrocatalysts that combine high conversion rates with precise control over reaction products. To overcome these limitations, we have developed a composite in which copper sulfide nanoflowers are grown directly on a reduced graphene oxide (rGO) matrix via a one‐pot synthesis. Detailed material analysis indicates that the presence of rGO promotes a crystallographic shift from the covellite (CuS) structure to a metastable digenite (Cu 7.2 S 4 ) phase. Spectroscopic investigations verify substantial electron transfer at the Cu 7.2 S 4 /rGO interface, leading to a modified electronic configuration around the copper sites. In electrochemical tests, the hybrid materials display markedly improved CO 2 reduction performance. The optimized Cu 7.2 S 4 /rGO‐10 catalyst delivers a Faradaic efficiency (FE) of 82% for formate at −0.5 V versus reversible hydrogen electrode, a significant improvement over the pure CuS benchmark (64%). Furthermore, tuning the rGO loading redirects the catalytic selectivity, with the Cu 7.2 S 4 /rGO‐40 variant yielding a 75% combined FE for liquid products, predominantly ethanol. These composites also demonstrate minimal hydrogen evolution and superior long‐term stability. This work underscores the multifunctional contribution of rGO in stabilizing unconventional material phases, enhancing electrical conductivity, and directing reaction mechanisms, thereby establishing a generalizable approach for fabricating high‐performance electrocatalysts for carbon‐neutral technologies.

From the perspective of scene communication theory, this study, based on scene communication theory, analyzed more than 200 short videos of 12 new farmers and found that the construction of content scenes, communication scenes, and consumption scenes has played a role in broadening sales channels, promoting employment, inheriting culture, and promoting the integration of agriculture, culture and tourism. But there are also problems such as content homogeneity, weak operational capacity, poor supply chain, and low conversion rate. Based on this, optimization paths are proposed from four dimensions: in-depth content development, enhanced digital literacy, improved supply chain, and policy empowerment, to provide references for the high-quality development of short-video sales by new farmers in Ganzi Prefecture and similar ethnic areas.