A New Sufficient & Necessary Condition for Testing Linear Separability Between Two Sets.

As a fundamental mathematical problem in the field of machine learning, the linear separability test still lacks a theoretically complete and computationally efficient method. This paper proposes and proves a sufficient and necessary condition for linear separability test based on a sphere model. The advantage of this test method is two-fold: (1) it provides not only a qualitative test of linear separability but also a quantitative analysis of the separability of linear separable instances; (2) it has low time cost and is more efficient than existing test methods. The proposed method is validated through a large number of experiments on benchmark datasets and artificial datasets, demonstrating both its correctness and efficiency.

Auto-Encoding Generative Adversarial Networks towards Mode Collapse Reduction and Feature Representation Enhancement.

Generative Adversarial Nets (GANs) are a kind of transformative deep learning framework that has been frequently applied to a large variety of applications related to the processing of images, video, speech, and text. However, GANs still suffer from drawbacks such as mode collapse and training instability. To address these challenges, this paper proposes an Auto-Encoding GAN, which is composed of a set of generators, a discriminator, an encoder, and a decoder. The set of generators is responsible for learning diverse modes, and the discriminator is used to distinguish between real samples and generated ones. The encoder maps generated and real samples to the embedding space to encode distinguishable features, and the decoder determines from which generator the generated samples come and from which mode the real samples come. They are jointly optimized in training to enhance the feature representation. Moreover, a clustering algorithm is employed to perceive the distribution of real and generated samples, and an algorithm for cluster center matching is accordingly constructed to maintain the consistency of the distribution, thus preventing multiple generators from covering a certain mode. Extensive experiments are conducted on two classes of datasets, and the results visually and quantitatively demonstrate the preferable capability of the proposed model for reducing mode collapse and enhancing feature representation.

Anti-tumor effect of PD-L1-targeting antagonistic aptamer-ASO delivery system with dual inhibitory function in immunotherapy.

Checkpoint inhibitor antibody therapy by blocking the interaction of surface programmed death-ligand 1(PD-L1) and programmed cell death protein 1(PD-1) has promising advantages in cancer immunotherapy. However, the response of many patients remains unsatisfactorily, suspected to be relevant to PD-L1 located in other cellular compartments and antibodies do not have access to the intracellular compartments. Herein, we identify a PD-L1-targeting DNA aptamer (PA9-1) with dual roles, including an antagonist and a delivery agent dependent on PD-L1 internalization. And we design the PD-L1-targeting antagonistic aptamer-ASO delivery system (PA9-1-ASO), with synergistic inhibitory PD-L1 activity involving the combination of blockade and silencing mechanisms. This chimera not only blocks PD-L1/PD-1 but also achieves targeted delivery of the conjugated ASO to reduce both surface PD-L1 and total PD-L1 expression. Compared with the single blockade, this chimera with the dual inhibitory function synergistically inhibits PD-L1 to amplify immunotherapeutic efficacy, providing a promising synergistic strategy for immunotherapy.

Metal-Organic Frameworks for Electrocatalytic Sensing of Hydrogen Peroxide.

The electrochemical detection of hydrogen peroxide (H2O2) has become more and more important in industrial production, daily life, biological process, green energy chemistry, and other fields (especially for the detection of low concentration of H2O2). Metal organic frameworks (MOFs) are promising candidates to replace the established H2O2 sensors based on precious metals or enzymes. This review summarizes recent advances in MOF-based H2O2 electrochemical sensors, including conductive MOFs, MOFs with chemical modifications, MOFs-composites, and MOF derivatives. Finally, the challenges and prospects for the optimization and design of H2O2 electrochemical sensors with ultra-low detection limit and long-life are presented.

Modeling human neurodevelopmental diseases with brain organoids.

Studying the etiology of human neurodevelopmental diseases has long been a challenging task due to the brain's complexity and its limited accessibility. Human pluripotent stem cells (hPSCs)-derived brain organoids are capable of recapitulating various features and functionalities of the human brain, allowing the investigation of intricate pathogenesis of developmental abnormalities. Over the past years, brain organoids have facilitated identifying disease-associated phenotypes and underlying mechanisms for human neurodevelopmental diseases. Integrating with more cutting-edge technologies, particularly gene editing, brain organoids further empower human disease modeling. Here, we review the latest progress in modeling human neurodevelopmental disorders with brain organoids.

Stable overexpression of mutated PTEN in Chinese hamster ovary cells enhances their performance and therapeutic antibody production.

Chinese hamster ovary (CHO) cells with a high viable cell density (VCD), resilience to culture stress, and the capacity to continuously express recombinant proteins are highly desirable. Phosphatase and tension homology deleted on chromosome ten (PTEN) functions as a key negative regulator of the PI3K/Akt signaling pathway, mediating cell growth and survival. Its oncogenic mutant endows cells with an enhanced proliferation rate and resistance to death. In this study, the role of oncogenic PTEN C124S or G129E on the performance of CHO-K1 and CHO-IgG cells was investigated. Our results showed that CHO-K1 cells stably expressing PTEN C124S or G129E exhibited enhanced proliferation, reduced apoptosis rate, and increased transient expression of therapeutic antibodies compared to the control cells. Moreover, the stable overexpression of PTEN C124S or G129E endowed CHO-IgG cells with higher cell viability, VCD, and antibody titers (yield increased by approximately 0.77-fold) in the fed-batch culture process and enhanced their performance in response to the addition of sodium lactate. Moreover, the engineering of mutated PTEN in CHO-IgG cells did not alter antibody quality. Collectively, our data suggest that mutated PTEN is a potential target for improving the manufacture of therapeutic antibodies.

Broadband infrared photoluminescence in silicon nanowires with high density stacking faults.

Making silicon an efficient light-emitting material is an important goal of silicon photonics. Here we report the observation of broadband sub-bandgap photoluminescence in silicon nanowires with a high density of stacking faults. The photoluminescence becomes stronger and exhibits a blue shift under higher laser powers. The super-linear dependence on excitation intensity indicates a strong competition between radiative and defect-related non-radiative channels, and the spectral blue shift is ascribed to the band filling effect in the heterostructures of wurtzite silicon and cubic silicon created by stacking faults.

Efficient solar water-splitting using a nanocrystalline CoO photocatalyst.

The generation of hydrogen from water using sunlight could potentially form the basis of a clean and renewable source of energy. Various water-splitting methods have been investigated previously, but the use of photocatalysts to split water into stoichiometric amounts of H2 and O2 (overall water splitting) without the use of external bias or sacrificial reagents is of particular interest because of its simplicity and potential low cost of operation. However, despite progress in the past decade, semiconductor water-splitting photocatalysts (such as (Ga1-xZnx)(N1-xOx)) do not exhibit good activity beyond 440 nm (refs 1,2,9) and water-splitting devices that can harvest visible light typically have a low solar-to-hydrogen efficiency of around 0.1%. Here we show that cobalt(II) oxide (CoO) nanoparticles can carry out overall water splitting with a solar-to-hydrogen efficiency of around 5%. The photocatalysts were synthesized from non-active CoO micropowders using two distinct methods (femtosecond laser ablation and mechanical ball milling), and the CoO nanoparticles that result can decompose pure water under visible-light irradiation without any co-catalysts or sacrificial reagents. Using electrochemical impedance spectroscopy, we show that the high photocatalytic activity of the nanoparticles arises from a significant shift in the position of the band edge of the material.

Anti-inflammatory activity of polysaccharide from Pholiota nameko.

Pholiota nameko polysaccharide (PNPS-1) has been isolated and purified by enzymatic hydrolysis, hot water extraction, ethanol precipitation, and ion-exchange and gel-filtration chromatography. The anti-inflammatory activity of PNPS-1 was evaluated in rodents using xylene-induced ear edema, egg albumin-, carrageenin-, and formaldehyde-induced paw edema, cotton pellet granuloma test, adhesion of peritoneal leukocytes in vitro, and ulcerogenic activity. The results showed that PNPS-1 (5 mg/ear) inhibited topical edema in the mouse ear and at 100, 200, and 400 mg/kg (intraperitoneally) it significantly suppressed the development of egg albumin-, carrageenin-, and formaldehyde-induced paw edema in the animals. PNPS-1 (100, 200, and 400 mg/kg, per oral) significantly inhibited the growth of granuloma tissues induced by subcutaneously implanted cotton pellets in rats by 10.96, 18.07, and 43.75%, respectively. PNPS-1 also inhibited spontaneous and phorbol-12-myristate-13-acetate-activated adhesion of peritoneal leukocytes in vitro. Further, both acute as well as chronic administration of PNPS-1 (100, 200, and 400 mg/kg, per oral) did not produce any gastric lesion in rats. In conclusion, these data indicated that PNPS-1 possesses significant anti-inflammatory activity suggesting its potential as an anti-inflammatory agent for use in the treatment of various inflammatory-related diseases.

Isolation, purification and antitumor activity of lipopolysaccharide from cow placenta.

Lipopolysaccharide from cow placenta (LPS-CP-2) has been isolated and purified by hot phenol-water extraction, enzyme hydrolysis, chloroform-petroleum ether method, ion-exchange and gel-filtration chromatography. Also, LPS-PS-2 was evaluated for antitumor activity against Ehrlich ascites carcinoma (EAC)-bearing Swiss albino mice. LPS-PS-2 caused significant (P<0.05) decrease in tumor volume, and viable cell count; and it prolonged the life span of EAC-tumor-bearing mice. Hematological profile indicates that LPS-CP-2 possessed protective action on the haemopoietic system. Further, administration of LPS-CP-2 reduced the tumor volume of both DLA and EAC cell lines in a dose-dependent way. The LPS-PS was found to be devoid of pyrogenic response in the rabbits. These results indicate that LPS-PS exhibited significant antitumor activity without pyrogenic response, suggesting its potential as antitumor agent.