Glucaric Acid Production from Miscanthus sacchariflorus via TEMPO-Mediated Oxidation with an Efficient Separation System.

In this study, production and isolation of glucaric acid from lignocellulosic biomass were performed via potassium cation-based TEMPO-mediated oxidation for the ease of glucaric acid isolation. To optimize the oxidation conditions, response surface methodology (RSM) was adopted using standard glucose as the raw material. Among the oxidation conditions, the dosage of oxidant and pH of reaction affected the glucaric acid production, and the optimum conditions were suggested by RSM analysis: 5 °C of reaction temperature, 4.23 equiv dosage of KClO per mole of glucose, and pH of 12. Furthermore, glucaric acid was produced from lignocellulosic biomass-derived enzymatic hydrolysate from Miscanthus under optimum conditions. The impurities such as xylose and lignin in enzymatic hydrolysate inhibited the efficiency of glucose oxidation. As a result, more oxidant was required to produce sufficient glucaric acid from the enzymatic hydrolysate compared to standard glucose. The produced glucaric acid was simply isolated by controlling the pH in the form of glucaric acid monopotassium salt, which showed lower solubility in water, and the purity of isolated glucaric acid was over 99%. The overall mass balance of feedstock to glucaric acid was analyzed, suggesting that 86.38% (w/w) glucaric acid could be produced from initial glucan in feedstock.

Agreeing to Stop: Reliable Latency-Adaptive Decision Making via Ensembles of Spiking Neural Networks.

Spiking neural networks (SNNs) are recurrent models that can leverage sparsity in input time series to efficiently carry out tasks such as classification. Additional efficiency gains can be obtained if decisions are taken as early as possible as a function of the complexity of the input time series. The decision on when to stop inference and produce a decision must rely on an estimate of the current accuracy of the decision. Prior work demonstrated the use of conformal prediction (CP) as a principled way to quantify uncertainty and support adaptive-latency decisions in SNNs. In this paper, we propose to enhance the uncertainty quantification capabilities of SNNs by implementing ensemble models for the purpose of improving the reliability of stopping decisions. Intuitively, an ensemble of multiple models can decide when to stop more reliably by selecting times at which most models agree that the current accuracy level is sufficient. The proposed method relies on different forms of information pooling from ensemble models and offers theoretical reliability guarantees. We specifically show that variational inference-based ensembles with p-variable pooling significantly reduce the average latency of state-of-the-art methods while maintaining reliability guarantees.

Immunoengineering can overcome the glycocalyx armour of cancer cells.

Cancer cell glycocalyx is a major line of defence against immune surveillance. However, how specific physical properties of the glycocalyx are regulated on a molecular level, contribute to immune evasion and may be overcome through immunoengineering must be resolved. Here we report how cancer-associated mucins and their glycosylation contribute to the nanoscale material thickness of the glycocalyx and consequently modulate the functional interactions with cytotoxic immune cells. Natural-killer-cell-mediated cytotoxicity is inversely correlated with the glycocalyx thickness of the target cells. Changes in glycocalyx thickness of approximately 10 nm can alter the susceptibility to immune cell attack. Enhanced stimulation of natural killer and T cells through equipment with chimeric antigen receptors can improve the cytotoxicity against mucin-bearing target cells. Alternatively, cytotoxicity can be enhanced through engineering effector cells to display glycocalyx-editing enzymes, including mucinases and sialidases. Together, our results motivate the development of immunoengineering strategies that overcome the glycocalyx armour of cancer cells.

Associations of cholecystectomy with metabolic health changes and incident cardiovascular disease: a retrospective cohort study.

Although some studies conducted about the risk of cholecystectomy and cardiovascular disease, there was a limit to explaining the relationship. We investigated the short-term and long-term relationship between cholecystectomy and cardiovascular disease, and evidence using the elements of the metabolic index as an intermediate step. It was a retrospective cohort study and we used the National Health Insurance Service database of South Korea between 2002 and 2015. Finally, 5,210 patients who underwent cholecystectomy and 49,457 at 1:10 age and gender-matched controls of subjects were collected. The main results was estimated by Multivariate Cox proportional hazard regression to calculate the hazard ratio (HR) with 95% confidence interval (CI) for risk of cardiovascular disease after cholecystectomy. Regarding short-term effects of cholecystectomy, increased risk of cardiovascular disease (aHR 1.35, 95% CI 1.15-1.58) and coronary heart disease (aHR 1.77, 95% CI 1.44-2.16) were similarly seen within 2 years of surgery. When analyzing the change in metabolic risk factors, cholecystectomy was associated with a change in systolic blood pressure (adjusted mean [aMean]: 1.51, 95% CI: [- 1.50 to - 4.51]), total cholesterol (aMean - 14.14, [- 20.33 to 7.95]) and body mass index (aMean - 0.13, [- 0.37 to 0.11]). Cholecystectomy patients had elevated risk of cardiovascular disease in the short-term, possibly due to the characteristics of the patient before surgery. The association of cholecystectomy and cardiovascular disease has decreased after 2 years in patients who underwent cholecystectomy, suggesting that because of improvement of metabolic health, cholecystectomy-associated elevation of cardiovascular disease risk may be ameliorated 2 years after cholecystectomy.

COLLAGEN MINERALIZATION DECREASES NK CELL-MEDIATED CYTOTOXICITY OF BREAST CANCER CELLS VIA INCREASED GLYCOCALYX THICKNESS.

Skeletal metastasis is common in patients with advanced breast cancer, and often caused by immune evasion of disseminated tumor cells (DTCs). In the skeleton, tumor cells not only disseminate to the bone marrow, but also to osteogenic niches in which they interact with newly mineralizing bone extracellular matrix (ECM). However, it remains unclear how mineralization of collagen type I, the primary component of bone ECM, regulates tumor-immune cell interactions. Here, we have utilized a combination of synthetic bone matrix models with controlled mineral content, nanoscale optical imaging, and flow cytometry to evaluate how collagen type I mineralization affects the biochemical and biophysical properties of the tumor cell glycocalyx, a dense layer of glycosylated proteins and lipids decorating their cell surface. Our results suggest that collagen mineralization upregulates mucin-type O-glycosylation and sialylation by tumor cells, which increased their glycocalyx thickness while enhancing resistance to attack by Natural Killer (NK) cells. These changes were functionally linked as treatment with a sialylation inhibitor decreased mineralization-dependent glycocalyx thickness and made tumor cells more susceptible to NK cell attack. Together, our results suggest that interference with glycocalyx sialylation may represent a therapeutic strategy to enhance cancer immunotherapies targeting bone-metastatic breast cancer.

Collagen Mineralization Decreases NK Cell-Mediated Cytotoxicity of Breast Cancer Cells via Increased Glycocalyx Thickness.

Skeletal metastasis is common in patients with advanced breast cancer and often caused by immune evasion of disseminated tumor cells (DTCs). In the skeleton, tumor cells not only disseminate to the bone marrow but also to osteogenic niches in which they interact with newly mineralizing bone extracellular matrix (ECM). However, it remains unclear how mineralization of collagen type I, the primary component of bone ECM, regulates tumor-immune cell interactions. Here, a combination of synthetic bone matrix models with controlled mineral content, nanoscale optical imaging, and flow cytometry are utilized to evaluate how collagen type I mineralization affects the biochemical and biophysical properties of the tumor cell glycocalyx, a dense layer of glycosylated proteins and lipids decorating their cell surface. These results suggest that collagen mineralization upregulates mucin-type O-glycosylation and sialylation by tumor cells, which increases their glycocalyx thickness while enhancing resistance to attack by natural killer (NK) cells. These changes are functionally linked as treatment with a sialylation inhibitor decreased mineralization-dependent glycocalyx thickness and made tumor cells more susceptible to NK cell attack. Together, these results suggest that interference with glycocalyx sialylation may represent a therapeutic strategy to enhance cancer immunotherapies targeting bone-metastatic breast cancer.

Depression and Subclinical Coronary Atherosclerosis in Adults Without Clinical Coronary Artery Disease.

BACKGROUND: The relationship between depression and subclinical coronary atherosclerosis in asymptomatic individuals is not clear. We evaluated this relationship in a Korean population. METHODS AND RESULTS: We analyzed 3920 individuals (mean age 54.7±7.9 years and 2603 men [66.4%]) with no history of coronary artery disease who voluntarily underwent coronary computed tomographic angiography and screening for depression using the Beck Depression Inventory as part of a general health examination. The degree and extent of subclinical coronary atherosclerosis were evaluated by coronary computed tomographic angiography, and ≥50% diameter stenosis was defined as significant. Participants were categorized into groups of those with or without depression using the Beck Depression Inventory scores ≥16 as a cutoff value. Of the study participants, 272 (6.9%) had a Beck Depression Inventory score of 16 or higher. After adjustment for cardiovascular risk factors, depression was not significantly associated with any coronary plaque (adjusted odds ratio [OR], 1.05 [95% CI, 0.78-1.41]; P=0.746), calcified plaque (OR, 0.95 [95% CI, 0.71-1.29]; P=0.758), noncalcified plaque (OR, 1.31 [95% CI, 0.79-2.17]; P=0.305), mixed plaque (OR, 1.16 [95% CI, 0.60-2.23]; P=0.659), or significant coronary artery stenosis (OR, 1.22 [95% CI, 0.73-2.03]; P=0.450). In the propensity score-matched population (n=1318) as well, none of the coronary artery disease measures of subclinical coronary atherosclerosis were statistically significantly associated with depression (all P>0.05). CONCLUSIONS: In this large cross-sectional study with asymptomatic individuals undergoing coronary computed tomographic angiography and Beck Depression Inventory evaluation, depression was not associated with an increased risk of subclinical coronary atherosclerosis.

Dietary butyrate and valerate glycerides impact diarrhea severity and immune response of weaned piglets under ETEC F4-ETEC F18 coinfection conditions.

Enterotoxigenic Escherichia coli (ETEC) causes post-weaning diarrhea in piglets, significantly impacting animal welfare and production efficiency. The two primary ETEC pathotypes associated with post-weaning diarrhea are ETEC F4 and ETEC F18. During the post-weaning period, piglets may be exposed to both ETEC F4 and ETEC F18. However, the effects of coinfection by both strains have not been studied. Short chain fatty acid feed additives, such as butyrate and valerate, are being investigated for their potential to improve animal performance and disease resistance. Therefore, this pilot experiment aimed to test the effects of butyrate glycerides or valerate glycerides on growth performance, diarrhea incidence, and immune responses of piglets under ETEC F4-ETEC F18 coinfection conditions. Twenty piglets were individually housed and assigned to one of the three dietary treatments immediately at weaning (21 to 24 d of age). The dietary treatments included control (basal diet formulation), control supplemented with 0.1% butyrate glycerides or 0.1% valerate glycerides. After a 7-d adaptation, all pigs were inoculated with ETEC F4 and ETEC F18 (0.5 × 109 CFU/1.5 mL dose for each strain) on three consecutive days. Pigs and feeders were weighed throughout the trial to measure growth performance. Fecal cultures were monitored for hemolytic coliforms, and blood samples were collected for whole blood and serum analysis. Pigs fed valerate glycerides tended (P = 0.095) to have higher final body weight compared with control. The overall severity of diarrhea was significantly (P < 0.05) lower in both treatment groups than control. Pigs fed valerate glycerides tended (P = 0.061) to have lower neutrophils and had significantly (P < 0.05) lower serum TNF-α on day 4 post-inoculation. This pilot experiment established an appropriate experimental dose for an ETEC F4-ETEC F18 coinfection disease model in weaned piglets. Results also suggest that butyrate glycerides and valerate glycerides alleviated diarrhea and regulated immune responses in piglets coinfected with ETEC F4 and ETEC F18.

Piglets suffer from post-weaning diarrhea associated with Enterotoxigenic Escherichia coli (ETEC) F4 and F18, two prevalent strains on swine farms globally. Short chain fatty acids (SCFAs), such as butyrate and valerate, are natural, organic compounds that could potentially promote intestinal health when used as dietary supplements. During the post-weaning period, piglets are vulnerable to simultaneous infection by ETEC F4 and F18. Therefore, this experiment aimed to develop an experimental disease model for coinfection with ETEC F4 and F18, employing a dose of 0.5 × 109 CFU/1.5 mL of each strain, administered over three consecutive days. In addition, the experiment evaluated treatment diets supplemented with 0.1% butyrate or valerate glycerides compared with the control diet. Results from this experiment revealed that the inoculation dose incited infection and diarrhea in piglets, implying its suitability for use in a disease challenge model. Moreover, the results indicated that the inclusion of butyrate and valerate glycerides to pig's diet reduced the severity of diarrhea. Furthermore, pigs fed SCFA glycerides exhibited lowered levels of inflammatory blood markers. In conclusion, the experimental dose induced diarrhea in piglets, and dietary supplementation of butyrate and valerate glycerides alleviated the severity of diarrhea while augmenting inflammatory status.

Electrostatic encoding of genome organization principles within single native nucleosomes.

The eukaryotic genome, first packed into nucleosomes of about 150 bp around the histone core, is organized into euchromatin and heterochromatin, corresponding to the A and B compartments, respectively. Here, we asked if individual nucleosomes in vivo know where to go. That is, do mono-nucleosomes by themselves contain A/B compartment information, associated with transcription activity, in their biophysical properties? We purified native mono-nucleosomes to high monodispersity and used physiological concentrations of biological polyamines to determine their condensability. The chromosomal regions known to partition into A compartments have low condensability and vice versa. In silico chromatin polymer simulations using condensability as the only input showed that biophysical information needed to form compartments is all contained in single native nucleosomes and no other factors are needed. Condensability is also strongly anticorrelated with gene expression, and especially so near the promoter region and in a cell type dependent manner. Therefore, individual nucleosomes in the promoter know whether the gene is on or off, and that information is contained in their biophysical properties. Comparison with genetic and epigenetic features suggest that nucleosome condensability is a very meaningful axis onto which to project the high dimensional cellular chromatin state. Analysis of condensability using various condensing agents including those that are protein-based suggests that genome organization principle encoded into individual nucleosomes is electrostatic in nature. Polyamine depletion in mouse T cells, by either knocking out ornithine decarboxylase (ODC) or inhibiting ODC, results in hyperpolarized condensability, suggesting that when cells cannot rely on polyamines to translate biophysical properties of nucleosomes to control gene expression and 3D genome organization, they accentuate condensability contrast, which may explain dysfunction known to occur with polyamine deficiency.

Biodegradation behavior of acetylated lignin added polylactic acid under thermophilic composting conditions.

Acetylated lignin (AL) can improve compatibility with commercial plastic polymers compared to existing lignin and can be used as an effective additive for eco-friendly biocomposites. For this reason, AL can be effectively incorporated into polylactic acid (PLA)-based biocomposites, but its biodegradation properties have not been investigated. In this study, biodegradation experiments were performed under mesophilic and thermophilic conditions to determine the effect of AL addition on the biodegradation characteristics of PLA-based biocomposites. As a result, the PLA-based biocomposite showed a faster biodegradation rate in a thermophilic composting environment, which is higher than the glass transition temperature of PLA, compared to a mesophilic environment. 16S rDNA sequencing results showed that differences in microbial communities depending on mesophilic and thermophilic environments strongly affected the biodegradation rate of lignin/PLA biocomposites. Importantly, the addition of AL can effectively delay the thermophilic biodegradation of PLA biocomposites. As a result of tracking the changes in physicochemical properties according to the biodegradation period in a thermophilic composting environment, the main biodegradation mechanism of AL/PLA biocomposite hydrolysis. It proceeded with cleavage of the PLA molecular chain, preferential biodegradation of the amorphous region, and additional biodegradation of the crystalline region. Above all, adding AL can be proposed as an effective additive because it can minimize the decline in the mechanical properties of PLA and delay the biodegradation rate more effectively compared to existing kraft lignin (KL).

Establish a machine learning based model for optimal casting conditions management of small and medium sized die casting manufacturers.

Die casting is a suitable process for producing complex and high precision parts, but it faces challenges in terms of quality degradation due to inevitable defects. The casting parameters play a significant role in quality, and in many cases, producers rely on their experience to manage these parameters. In order to address this, domestic small and medium sized die casting companies have established smart factories (MES) and collected data. This study aims to utilize this data to construct a machine learning based optimal casting parameter model to enhance quality. During the model development process, distinct important features were identified for each company. This indicates the necessity of deriving tailored models for each site, aligning with the make to order (MTO) environment, rather than a generalized model.

Bi-directional nucleosome sliding by the Chd1 chromatin remodeler integrates intrinsic sequence-dependent and ATP-dependent nucleosome positioning.

Chromatin remodelers use a helicase-type ATPase motor to shift DNA around the histone core. Although not directly reading out the DNA sequence, some chromatin remodelers exhibit a sequence-dependent bias in nucleosome positioning, which presumably reflects properties of the DNA duplex. Here, we show how nucleosome positioning by the Chd1 remodeler is influenced by local DNA perturbations throughout the nucleosome footprint. Using site-specific DNA cleavage coupled with next-generation sequencing, we show that nucleosomes shifted by Chd1 can preferentially localize DNA perturbations - poly(dA:dT) tracts, DNA mismatches, and single-nucleotide insertions - about a helical turn outside the Chd1 motor domain binding site, super helix location 2 (SHL2). This phenomenon occurs with both the Widom 601 positioning sequence and the natural +1 nucleosome sequence from the Saccharomyces cerevisiae SWH1 gene. Our modeling indicates that localization of DNA perturbations about a helical turn outward from SHL2 results from back-and-forth sliding due to remodeler action on both sides of the nucleosome. Our results also show that barrier effects from DNA perturbations can be extended by the strong phasing of nucleosome positioning sequences.

Photosensitizing deep-seated cancer cells with photoprotein-conjugated upconversion nanoparticles.

To resolve the problem of target specificity and light transmission to deep-seated tissues in photodynamic therapy (PDT), we report a cancer cell-targeted photosensitizer using photoprotein-conjugated upconversion nanoparticles (UCNPs) with high target specificity and efficient light transmission to deep tissues. Core-shell UCNPs with low internal energy back transfer were conjugated with recombinant proteins that consists of a photosensitizer (KillerRed; KR) and a cancer cell-targeted lead peptide (LP). Under near infrared (NIR)-irradiating condition, the UCNP-KR-LP generated superoxide anion radicals as reactive oxygen species via NIR-to-green light conversion and exhibited excellent specificity to target cancer cells through receptor-mediated cell adhesion. Consequently, this photosensitizing process facilitated rapid cell death in cancer cell lines (MCF-7, MDA-MB-231, and U-87MG) overexpressing integrin beta 1 (ITGB1) receptors but not in a cell line (SK-BR-3) with reduced ITGB1 expression and a non-invasive normal breast cell line (MCF-10A). In contrast to green light irradiation, NIR light irradiation exhibited significant PDT efficacy in cancer cells located beneath porcine skin tissues up to a depth of 10 mm, as well as in vivo tumor xenograft mouse models. This finding suggests that the designed nanocomposite is useful for sensing and targeting various deep-seated tumors.

The microenvironment dictates glycocalyx construction and immune surveillance.

Efforts to identify anti-cancer therapeutics and understand tumor-immune interactions are built with in vitro models that do not match the microenvironmental characteristics of human tissues. Using in vitro models which mimic the physical properties of healthy or cancerous tissues and a physiologically relevant culture medium, we demonstrate that the chemical and physical properties of the microenvironment regulate the composition and topology of the glycocalyx. Remarkably, we find that cancer and age-related changes in the physical properties of the microenvironment are sufficient to adjust immune surveillance via the topology of the glycocalyx, a previously unknown phenomenon observable only with a physiologically relevant culture medium.

Exploration of Potential Breath Biomarkers of Chronic Kidney Disease through Thermal Desorption-Gas Chromatography/Mass Spectrometry.

Breath volatile organic compound (VOC) analysis is a non-invasive tool for assessing health status; the compositional profile of these compounds in the breath of patients with chronic kidney disease is believed to change with decreasing renal function. We aimed to identify breath VOCs for recognizing patients with chronic kidney disease. Using thermal desorption-gas chromatography/mass spectrometry, untargeted analysis of breath markers was performed using breath samples of healthy controls (n = 18) versus non-dialysis (n = 21) and hemodialysis (n = 12) patients with chronic kidney disease in this cross-sectional study. A total of 303 VOCs alongside 12 clinical variables were used to determine the breath VOC profile. Metabolomic analysis revealed that age, systolic blood pressure, and fifty-eight breath VOCs differed significantly between the chronic kidney disease group (non-dialysis + hemodialysis) and healthy controls. Thirty-six VOCs and two clinical variables that showed significant associations with chronic kidney disease in the univariate analysis were further analyzed. Different spectra of breath volatile organic compounds between the control and chronic kidney disease groups were obtained. A multivariate model incorporating age, 2-methyl-pentane, and cyclohexanone showed high performance (accuracy, 86%) in identifying patients with chronic kidney disease with odds ratios of 0.18 (95% CI, 0.07-2.49, p = 0.013); 2.10 (0.94-2.24, p = 0.025); and 2.31 (0.88-2.64, p = 0.008), respectively. Hence, this study showed that renal dysfunction induces a characteristic profile of breath VOCs that can be used as non-invasive potential biomarkers in screening tests for CKD.

Differential Impact of Degree of Hypertension on Subclinical Coronary Atherosclerosis in Asymptomatic Subjects With and Without Diabetes Mellitus.

This study sought to evaluate the association between the degree of hypertension and subclinical coronary atherosclerosis in asymptomatic subjects with and without diabetes mellitus (DM). We retrospectively analyzed 7,352 asymptomatic subjects (mean age 52.8 ± 7.8 years; 4,689 [63.8%] men) with no history of coronary artery disease who voluntarily underwent coronary computed tomography angiography as part of a general health examination. The classification of hypertension was adapted from the American College of Cardiology and American Heart Association 2017 guideline. Subclinical coronary atherosclerosis was defined as the presence of coronary plaque by coronary computed tomography angiography. In subjects without DM (n = 6,598), after the adjustment for cardiovascular risk factors, subclinical coronary atherosclerosis was significantly associated with both stage 1 hypertension (adjusted odds ratio [aOR] 1.356; 95% confidence interval [CI], 1.167 to 1.575; p <0.001) and stage 2 hypertension (aOR, 1.614; 95% CI, 1.329 to 1.961; p <0.001) groups compared with the normal group. In contrast, in subjects with DM (n = 754), there was no statistical difference in the aOR of the stage 1 hypertension group for the presence of coronary plaque (aOR, 1.449; 95% CI, 0.982 to 2.136; p = 0.061). However, the stage 2 hypertension group had a significant association with subclinical coronary atherosclerosis (aOR, 2.067; 95% CI, 1.287 to 3.322; p = 0.003). In subjects without DM, both stages 1 and 2 hypertension were associated with subclinical coronary atherosclerosis. However, in subjects with DM, stage 2 hypertension was only associated with an increased risk of subclinical coronary atherosclerosis.

Impact of Trimetazidine on the Incident Heart Failure After Coronary Artery Revascularization.

ABSTRACT: Abnormal myocardial metabolism is a common pathophysiological process underlying ischemic heart disease and heart failure (HF). Trimetazidine is an antianginal agent with a unique mechanism of action that regulates myocardial energy metabolism and might have a beneficial effect in preventing HF in patients undergoing myocardial revascularization. We aimed to evaluate the potential benefit of trimetazidine in preventing incident hospitalization for HF after myocardial revascularization. Using the common data model, we identified patients without prior HF undergoing myocardial revascularization from 8 hospital databases in Korea. To compare clinical outcomes using trimetazidine, database-level hazard ratios (HRs) were estimated using large-scale propensity score matching for each database and pooled using a random-effects model. The primary outcome was incident hospitalization for HF. The secondary outcome of interest was major adverse cardiac events (MACEs). After propensity score matching, 6724 and 11,211 patients were allocated to trimetazidine new-users and nonusers, respectively. There was no significant difference in the incidence of hospitalization for HF between the 2 groups (HR: 1.08, 95% confidence interval [CI], 0.88-1.31; P = 0.46). The risk of MACE also did not differ between the 2 groups (HR: 1.07, 95% CI, 0.98-1.16; P = 0.15). In conclusion, the use of trimetazidine did not reduce the risk of hospitalization for HF or MACE in patients undergoing myocardial revascularization. Therefore, the role of trimetazidine in contemporary clinical practice cannot be expanded beyond its current role as an add-on treatment for symptomatic angina.

Robust PAC m : Training Ensemble Models Under Misspecification and Outliers.

Standard Bayesian learning is known to have suboptimal generalization capabilities under misspecification and in the presence of outliers. Probably approximately correct (PAC)-Bayes theory demonstrates that the free energy criterion minimized by Bayesian learning is a bound on the generalization error for Gibbs predictors (i.e., for single models drawn at random from the posterior) under the assumption of sampling distributions uncontaminated by outliers. This viewpoint provides a justification for the limitations of Bayesian learning when the model is misspecified, requiring ensembling, and when data are affected by outliers. In recent work, PAC-Bayes bounds-referred to as PAC m -were derived to introduce free energy metrics that account for the performance of ensemble predictors, obtaining enhanced performance under misspecification. This work presents a novel robust free energy criterion that combines the generalized logarithm score function with PAC m ensemble bounds. The proposed free energy training criterion produces predictive distributions that are able to concurrently counteract the detrimental effects of misspecification-with respect to both likelihood and prior distribution-and outliers.

pH-sensitive cellulose/chitin nanofibrillar hydrogel for dye pollutant removal.

In this study, a pH-sensitive smart hydrogel was successfully prepared by combining a polyelectrolyte complex using biopolymeric nanofibrils. By adding a green citric acid cross-linking agent to the formed chitin and cellulose-derived nanofibrillar polyelectrolytic complex, a hydrogel with excellent structural stability could be prepared even in a water environment, and all processes were conducted in an aqueous system. The prepared biopolymeric nanofibrillar hydrogel not only enables rapid conversion of swelling degree and surface charge according to pH but can also effectively remove ionic contaminants. The ionic dye removal capacity was 372.0 mg/g for anionic AO and 140.5 mg/g for cationic MB. The surface charge conversion ability according to pH could be easily applied to the desorption of the removed contaminants, and as a result, it showed an excellent contaminant removal efficiency of 95.1 % or more even in the repeated reuse process 5 times. Overall, the eco-friendly biopolymeric nanofibrillar pH-sensitive hydrogel shows potential for complex wastewater treatment and long-term use.