LF3PFL: A Practical Privacy-Preserving Federated Learning Algorithm Based on Local Federalization Scheme.

In the realm of federated learning (FL), the exchange of model data may inadvertently expose sensitive information of participants, leading to significant privacy concerns. Existing FL privacy-preserving techniques, such as differential privacy (DP) and secure multi-party computing (SMC), though offering viable solutions, face practical challenges including reduced performance and complex implementations. To overcome these hurdles, we propose a novel and pragmatic approach to privacy preservation in FL by employing localized federated updates (LF3PFL) aimed at enhancing the protection of participant data. Furthermore, this research refines the approach by incorporating cross-entropy optimization, carefully fine-tuning measurement, and improving information loss during the model training phase to enhance both model efficacy and data confidentiality. Our approach is theoretically supported and empirically validated through extensive simulations on three public datasets: CIFAR-10, Shakespeare, and MNIST. We evaluate its effectiveness by comparing training accuracy and privacy protection against state-of-the-art techniques. Our experiments, which involve five distinct local models (Simple-CNN, ModerateCNN, Lenet, VGG9, and Resnet18), provide a comprehensive assessment across a variety of scenarios. The results clearly demonstrate that LF3PFL not only maintains competitive training accuracies but also significantly improves privacy preservation, surpassing existing methods in practical applications. This balance between privacy and performance underscores the potential of localized federated updates as a key component in future FL privacy strategies, offering a scalable and effective solution to one of the most pressing challenges in FL.

Sonochemical synthesis of CoNi layered double hydroxide as a cathode material for assembling high performance hybrid supercapacitor.

Fabricating battery-type electrode materials with large specific surface area and mesopores is an efficient method for enhancing the electrochemical performance of supercapacitors. This method may provide more active sites for Faradic reactions and shorten the ion-diffusion paths. In this study, the CoNi layered double hydroxides (LDHs) with the morphology of nanoflowers and nanoflakes were prepared in solutions with pH values of 7.5 (CoNi LDH-7.5) and 8.5 (CoNi LDH-8.5) via a simple sonochemical approach. These CoNi LDHs possessed large specific surface areas and favourable electrochemical properties. The CoNi LDH-7.5 delivered a specific capacity of 740.8C/g at a current density of 1 A/g, surpassing CoNi LDH-8.5 with 668.1C/g. The hybrid supercapacitor (HSC) was assembled with activated carbon as the anode and CoNi LDH as the cathode to assess its practical application potential in the field of electrochemical energy storage. The CoNi LDH-7.5//AC HSC achieved the highest energy density of 35.6 W h kg-1 at a power density of 781.1 W kg-1. In addition, both HSCs exhibited little capacity decay over 5,000 cycles at a high current load of 8 A/g. These electrochemical properties of CoNi LDHs make them promising candidates for battery-type electrode materials. The current sonochemical method is simple and can be applied to the preparation of other LDHs-based electrode materials with favourable electrochemical performance.

An Optimization Method for Non-IID Federated Learning Based on Deep Reinforcement Learning.

Federated learning (FL) is a distributed machine learning paradigm that enables a large number of clients to collaboratively train models without sharing data. However, when the private dataset between clients is not independent and identically distributed (non-IID), the local training objective is inconsistent with the global training objective, which possibly causes the convergence speed of FL to slow down, or even not converge. In this paper, we design a novel FL framework based on deep reinforcement learning (DRL), named FedRLCS. In FedRLCS, we primarily improved the greedy strategy and action space of the double DQN (DDQN) algorithm, enabling the server to select the optimal subset of clients from a non-IID dataset to participate in training, thereby accelerating model convergence and reaching the target accuracy in fewer communication epochs. In simulation experiments, we partition multiple datasets with different strategies to simulate non-IID on local clients. We adopt four models (LeNet-5, MobileNetV2, ResNet-18, ResNet-34) on the four datasets (CIFAR-10, CIFAR-100, NICO, Tiny ImageNet), respectively, and conduct comparative experiments with five state-of-the-art non-IID FL methods. Experimental results show that FedRLCS reduces the number of communication rounds required by 10-70% with the same target accuracy without increasing the computation and storage costs for all clients.

Discovery of a novel, highly potent and orally bioavailable pyrrolidinone indole series of irreversible Myeloperoxidase (MPO) inhibitors.

Myeloperoxidase (MPO) is a heme-containing peroxidase from phagocytic cells, which plays an important role in the innate immune response. The primary anti-microbial function of MPO is achieved by catalyzing the oxidation of halides by hydrogen peroxide (H2O2). Upon activation of phagocytes, MPO activity is detectable in both phagosomes and extracellularly, where it can remain or transcytose into interstitial compartments. Activated MPO leads to oxidative stress and tissue damage in many inflammatory states, including cardiovascular disease. Starting from a low molecular weight (LMW) high throughput screening (HTS) hit, here we report the discovery of a novel pyrrolidinone indole (IN-4) as a highly potent MPO inhibitor. This compound displays similar in vitro potency across peroxidation, plasma and NETosis assays. In a dilution/dialysis study, <5% of the original MPO activity was detected post-incubation of MPO with IN-4, suggesting irreversible enzyme inhibition. A fast MPO inactivation rate (kinact/Ki) and low partition ratio (k3/k4) make IN-4 kinetic properties attractive for an MPO inhibitor. This compound also displays significant selectivity over the closely related thyroid peroxidase (TPO), and is selective for extracellular MPO over intracellular (neutrophil) MPO. Moreover, IN-4 shows good exposure, low clearance and high oral bioavailability in mice, rats and dogs. The high in vitro MPO activity and high oral exposure observed with IN-4 result in a dose-dependent inhibition of MPO activity in three mouse models of inflammation. In conclusion, IN-4 is a novel, potent, mechanism-based and selective MPO inhibitor, which may be used as superior therapeutic agent to treat multiple inflammatory conditions, including cardiovascular disease.

A Refined Blood Pressure Estimation Model Based on Single Channel Photoplethysmography.

This study proposed a refined BP prediction strategy that using single-channel photoplethysmography (PPG) signals to stratify populations by cardiovascular status before BP estimation. Combining demographic characteristics (age, gender) and pulse wave morphological features, the random forest was applied to screen two kinds of typical cardiovascular diseases (CVDs), with an accuracy of 92.2%. A deep learning model (BiLSTM-At) was proposed to estimate the long-term BP trend for different CVD groups. Transfer learning technique was used for personalized modeling to reduce computational complexity while improving performance. The method was validated on 255 patients with different CVDs. The mean absolute errors (MAEs) of systolic blood pressure (SBP) and diastolic blood pressure (DBP) estimation were 2.815 mmHg and 1.876 mmHg for normal subjects, 3.024 mmHg and 1.334 mmHg for AF subjects, and 4.444 mmHg and 2.549 mmHg for CA subjects. The results met the American Association for the Advancement of Medical Instrumentations (AAMI) and British Hypertension Society (BHS) Class A criteria. This indicated that our strategy has good performance and can realize long-term monitoring of BP through a small batch samples, with the potential to implement real-time monitoring in healthy devices.

Porous MgCo2O4 nanoflakes serve as electrode materials for hybrid supercapacitors with excellent performance.

In this work, MgCo2O4 microspheres (MgCo2O4 MSs) and MgCo2O4 nanoflakes (MgCo2O4 NFs) were prepared by one-step and two-step synthetic method, respectively, and combined with a post annealing treatment. Both MSs and NFs electrode materials possessed porous structure and large specific surface areas. The electrochemical properties were evaluated using three-electrode as well as two-electrode systems. The MgCo2O4 NFs delivered a specific capacity of 375.5C g-1 at 1 A g-1 together with a high rate performance (74.9%) at 10 A g-1, while the MgCo2O4 MSs exhibited 276.3C g-1 at the current density of 1 A g-1. A hybrid supercapacitor (HSC) device was assembled with a cathode made from MgCo2O4 and an anode made from activated carbon (AC) for evaluation of real applications, and it was able to run over a high voltage window (1.75 V). This MgCo2O4 NFs//AC HSC delivered a high energy density (Ed, 35.4 W h kg-1) at 950.6 W kg-1, and at the highest power density (Pd) of 8905.0 W kg-1, it could still hold 25.8 W h kg-1. On the other hand, the MgCo2O4 MSs//AC HSC device exhibited an Ed of 32.4 W h kg-1 at a Pd of 1048.0 W kg-1. Both HSCs exhibited good long-term cycling stability due to no capacity decay over 6000 cycles at 6 A g-1. The excellent electrochemical performance demonstrates that these MgCo2O4 electrode materials, especially the MgCo2O4 NFs, have great application potential for electrochemical energy storage. This synthesis method is simple and is possibly to be applied in synthesizing other transition metal oxides (TMOs)-based electrode materials with large surface area and outstanding electrochemical performance.

Nanosheet-assembled porous MnCo2O4.5 microflowers as electrode material for hybrid supercapacitors and lithium-ion batteries.

Herein, the MnCo2O4.5 microflowers (MFs) assembled by two-dimensional (2D) porous nanosheets were prepared through an initial solvothermal reaction with a subsequent annealing process. In this architecture, many interconnected 2D thin nanosheets were self-assembled together to form a 3D hierarchical MF with plenty of open channels. Such structure endows these MnCo2O4.5 MFs with large specific surface area of 156.85 m2/g for energy storage and provides rich ion diffusion pathways for ion transportation, thus the as-prepared MFs can exhibit good overall electrochemical performance in both hybrid supercapacitor (HSC) and lithium-ion battery (LIB). For the utilization in supercapacitor, the MFs deliver a specific capacity of 287.02 C/g at 1 A/g as well as a rate capability with 73.3 % capacity retention at 8 A/g. The energy density of the HSC assembled by MFs and activated carbon can reach up to 30.33 W h kg-1 at 959.35 W kg-1. When applied as the anode for Li-ion battery, a specific capacity of 1340.8 mA h g-1 at 0.1 A/g and cycling performance with low capacity loss of 0.73 mAh/g per cycle after 200 cycles at 0.5 A/g can be achieved. This work uncovers a repeatable and facile synthetic strategy to prepare transition metal oxides with large specific surface area and good overall electrochemical property.

Discovery of a novel indole pharmacophore for the irreversible inhibition of myeloperoxidase (MPO).

Myeloperoxidase (MPO) activity and subsequent generation of hypochlorous acid has been associated with the killing of host-invading microorganisms (e.g. bacteria, viruses, and fungi). However, during oxidative stress, high MPO activity can damage host tissue and is linked to several chronic inflammatory conditions. Herein, we describe the development of a novel biaryl, indole-pyrazole series of irreversible mechanism-based inhibitors of MPO. Derived from an indole-containing high-throughput screen hit, optimization efforts resulted in potent and selective 6-substituted indoles with good oral bioavailability and in vivo activity.

Discovery of 1-((6-Aminopyridin-3-yl)Methyl)-3-(4-Bromophenyl)Urea as a Potent, Irreversible Myeloperoxidase Inhibitor.

Myeloperoxidase (MPO) is a leukocyte-derived redox enzyme that has been linked to oxidative stress and damage in many inflammatory states, including cardiovascular disease. We have discovered aminopyridines that are potent mechanism-based inhibitors of MPO, with significant selectivity over the closely related thyroid peroxidase. 1-((6-Aminopyridin-3-yl)methyl)-3-(4-bromophenyl)urea (Aminopyridine 2) inhibited MPO in human plasma and blocked MPO-dependent vasomotor dysfunction ex vivo in rat aortic rings. Aminopyridine 2 also showed high oral bioavailability and inhibited MPO activity in vivo in a mouse model of peritonitis. Aminopyridine 2 could effectively be administered as a food admixture, making it an important tool for assessing the relative importance of MPO in preclinical models of chronic inflammatory disease.

The DPP-4 inhibitor vildagliptin increases pancreatic beta cell mass in neonatal rats.

The present study addressed the effect of the dipeptidyl peptidase-4 (DPP-4) inhibitor vildagliptin ((1-[[(3-hydroxy-1-adamantyl) amino] acetyl]-2-cyano-(S)-pyrrolidine), LAF237) on pancreatic beta cell mass in neonatal rats. Newborn rats were treated orally with vildagliptin (60 mg/kg) or vehicle once daily for 19 days starting from postnatal day 2. Pancreatic immunohistochemistry and morphometric analysis were performed to evaluate changes in beta cell mass, cell apoptosis (Apoptag stain) and replication (5'-Bromo-2'-deoxyuridine (BrdU)-incorporation) on days 7, 21, and 33. On day 7, an eight-fold increase in BrdU-positive pancreatic beta cells and a 71% decrease in Apoptag-positive cells were observed. On day 21, vildagliptin produced a two-fold increase in pancreatic beta cell mass compared to placebo (0.06±0.01 mg vs 0.11±0.02 mg, P<0.05). Beta cell mass remained elevated (90%, 0.09±0.02 mg vs 0.16±0.03 mg, P<0.05) on day 33, twelve days after discontinuing vildagliptin treatment. These data show that the DPP-4 inhibitor vildagliptin increased pancreatic beta cell mass through enhanced beta cell replication and reduced apoptosis. The increased beta cell mass was sustained for 12 days after vildagliptin washout. This study demonstrates that DPP-4 inhibitors can elicit beneficial effects on beta cell turnover that could help to prevent or retard the progression of type 2 diabetes.

SIRT4 regulates fatty acid oxidation and mitochondrial gene expression in liver and muscle cells.

SIRT4, a member of the sirtuin family, has been implicated in the regulation of insulin secretion by modulation of glutamate dehydrogenase. However, the role of this enzyme in the regulation of metabolism in other tissues is unknown. In this study we investigated whether depletion of SIRT4 would enhance liver and muscle metabolic functions. To do this SIRT4 was knocked down using an adenoviral shRNA in mouse primary hepatocytes and myotubes. We observed a significant increase in gene expression of mitochondrial and fatty acid metabolism enzymes in hepatocytes with reduced SIRT4 levels. SIRT4 knockdown also increased SIRT1 mRNA and protein levels both in vitro and in vivo. In agreement with the increased fatty acid oxidation (FAO) gene expression, we showed a significant increase in FAO in SIRT4 knockdown primary hepatocytes compared with control, and this effect was dependent on SIRT1. In primary myotubes, knockdown of SIRT4 resulted in increased FAO, cellular respiration, and pAMPK levels. When SIRT4 was knocked down in vivo by tail vein injection of a shRNA adenovirus, we observed a significant increase in hepatic mitochondrial and FAO gene expression consistent with the findings in primary hepatocytes. Taken together these findings demonstrate that SIRT4 inhibition increases fat oxidative capacity in liver and mitochondrial function in muscle, which might provide therapeutic benefits for diseases associated with ectopic lipid storage such as type 2 diabetes.

Cevoglitazar, a novel peroxisome proliferator-activated receptor-alpha/gamma dual agonist, potently reduces food intake and body weight in obese mice and cynomolgus monkeys.

Cevoglitazar is a dual agonist for the peroxisome proliferator-activated receptor (PPAR)-alpha and -gamma subtypes. Dual activation of PPARalpha and -gamma is a therapeutic approach in development for the treatment of type 2 diabetes mellitus and diabetic dyslipidemia. In this report, we show that, in addition to improving insulin sensitivity and lipid metabolism like other dual PPAR agonists, cevoglitazar also elicits beneficial effects on energy homeostasis in two animal models of obesity. In leptin-deficient ob/ob mice, administration of cevoglitazar at 0.5, 1, or 2 mg/kg for 18 d led to acute and sustained, dose-dependent reduction of food intake and body weight. Furthermore, plasma levels of glucose and insulin were normalized after 7 d of cevoglitazar treatment at 0.5 mg/kg. Plasma levels of free fatty acids and triglycerides were dose-dependently reduced. In obese and insulin-resistant cynomolgus monkeys, treatment with cevoglitazar at 50 and 500 mug/kg for 4 wk lowered food intake and body weight in a dose-dependent manner. In these animals, cevoglitazar also reduced fasting plasma insulin and, at the highest dose, reduced hemoglobin A1c levels by 0.4%. These preclinical results demonstrate that cevoglitazar holds promise for the treatment of diabetes and obesity-related disorders because of its unique beneficial effect on energy balance in addition to improving glycemic and metabolic control.