PLDP-FL: Federated Learning with Personalized Local Differential Privacy.

As a popular machine learning method, federated learning (FL) can effectively solve the issues of data silos and data privacy. However, traditional federated learning schemes cannot provide sufficient privacy protection. Furthermore, most secure federated learning schemes based on local differential privacy (LDP) ignore an important issue: they do not consider each client's differentiated privacy requirements. This paper introduces a perturbation algorithm (PDPM) that satisfies personalized local differential privacy (PLDP), resolving the issue of inadequate or excessive privacy protection for some participants due to the same privacy budget set for all clients. The algorithm enables clients to adjust the privacy parameters according to the sensitivity of their data, thus allowing the scheme to provide personalized privacy protection. To ensure the privacy of the scheme, we have conducted a strict privacy proof and simulated the scheme on both synthetic and real data sets. Experiments have demonstrated that our scheme is successful in producing high-quality models and fulfilling the demands of personalized privacy protection.

Privacy-Preserving Multi-Class Support Vector Machine Model on Medical Diagnosis.

With the rapid development of machine learning in the medical cloud system, cloud-assisted medical computing provides a concrete platform for remote rapid medical diagnosis services. Support vector machine (SVM), as one of the important algorithms of machine learning, has been widely used in the field of medical diagnosis for its high classification accuracy and efficiency. In some existing schemes, healthcare providers train diagnostic models with SVM algorithms and provide online diagnostic services to doctors. Doctors send the patient's case report to the diagnostic models to obtain the results and assist in clinical diagnosis. However, case report involves patients' privacy, and patients do not want their sensitive information to be leaked. Therefore, the protection of patient's privacy has become an important research direction in the field of online medical diagnosis. In this paper, we propose a privacy-preserving medical diagnosis scheme based on multi-class SVMs. The scheme is based on the distributed two trapdoors public key cryptosystem (DT-PKC) and Boneh-Goh-Nissim (BGN) cryptosystem. We design a secure computing protocol to compute the core process of the SVM classification algorithm. Our scheme can deal with both linearly separable data and nonlinear data while protecting the privacy of user data and support vectors. The results show that our scheme is secure, reliable, scalable with high accuracy.

A Single-Chain Variable Fragment Antibody Inhibits Aggregation of Phosphorylated Tau and Ameliorates Tau Toxicity in vitro and in vivo.

BACKGROUND: Alzheimer's disease (AD) is a common cause of dementia among elderly people. Hyperphosphorylation and aggregation of tau correlates with the clinical progression of AD; therefore, therapies targeting the aggregation of tau may have potential applications for anti-AD drug development. Several inhibitors of tau aggregation, including small molecules and antibodies, have been found to decrease the aggregation of tau and the corresponding pathology. OBJECTIVE: To screen one kind of single-chain variable fragment (scFv) antibody which could inhibit the aggregation of tau and ameliorate its cytotoxicity. METHODS/RESULTS: Using phosphorylated tau (pTau) as an antigen, we obtained a scFv antibody via the screening of a high-capacity phage antibody library. Biochemical analysis revealed that this scFv antibody (scFv T1) had a strong ability to inhibit pTau aggregation both in dilute solutions and under conditions of macromolecular crowding. ScFv T1 could also depolymerize preformed pTau aggregates in vitro. Furthermore, scFv T1 was found to be able to inhibit the cytotoxicity of extracellular pTau aggregates and ameliorate tau-mediated toxicity when coexpressed with a hTauR406W mutant in the eye of transgenic Drosophila flies. CONCLUSION: This scFv T1 antibody may be a potential new therapeutic agent against AD. Our methods can be used to develop novel strategies against protein aggregation for the treatment of neurodegenerative diseases.

Gut microbiota from colorectal cancer patients enhances the progression of intestinal adenoma in Apcmin/+ mice.

BACKGROUND: Accumulating evidence points to a close relationship between gut dysbiosis and colorectal cancer (CRC). As >90% of CRC develop from adenoma, we aimed to investigate the crucial role of imbalanced gut microbiota on the progression of intestinal adenoma. METHODS: The Apcmin/+ mice gavage with phosphate-buffered saline (PBS), feces from healthy controls or CRC patients after antibiotic cocktails. The intestinal tissues were isolated for histopathology, western blotting, and RNA-seq. The microbiota of feces and short-chain fatty acids (SCFAs) were analysed by 16S rDNA Amplicon Sequencing and gas chromatography. FINDINGS: The Apcmin/+mice gavaged by feces from CRC patients had more intestinal tumours compared with those fed with feces from healthy controls or PBS. Administration of feces from CRC patients increased tumour proliferation and decreased apoptosis in tumour cells, accompanied by impairment of gut barrier function and up-regulation the pro-inflammatory cytokines profile. The up-regulated the expression of ß-catenin and cyclinD1 further indicating the activation of Wnt signalling pathway. The abundance of pathogenic bacteria was increased after FMT, while producing SCFAs bacteria and SCFAs production were decreased. INTERPRETATION: Gut microbiota of CRC patients disrupted intestinal barrier, induced low-grade inflammation and dysbiosis. The altered gut microbiota enhanced the progression of intestinal adenomas in Apcmin/+mice, suggesting that a new strategy to target gut microbiota against CRC could be noted. FUND: The study was supported by the National Natural Science Foundation of China, Tianjin Research Programme of Application Foundation and Advanced Technology of China, and China Postdoctoral Science Foundation.

Organo-Catalyzed Regio- and Geometry-Specific Construction of ß-Hydroxyl-α-vinyl Carboxylic Esters: Substrate Scope, Mechanistic Insights, and Applications.

A green protocol has been developed for the synthesis of ß-hydroxyl-α-vinyl carboxylic esters using aldehydes and α,ß-unsaturated esters bearing an activated γ proton as starting materials under Morita-Baylis-Hillman (MBH) reaction conditions. Diverse ß-hydroxyl-α-vinyl carboxylic esters have been synthesized regiospecifically in moderate to good yields with only E geometric selectivity. Other remarkable features include atom efficiency, environmental benignancy, and mild reaction conditions. Furthermore, the reaction products could be readily converted into tetrahydrofuran, dihydrofuran, and furan derivatives.

Tudor-staphylococcal nuclease regulates the expression and biological function of alkylglycerone phosphate synthase via nuclear factor-κB and microRNA-127 in human glioma U87MG cells.

Glioma is one of the malignant tumor types detrimental to human health; therefore, it is important to find novel targets and therapeutics for this tumor. The downregulated expression of Tudor-staphylococcal nuclease (SN) and alkylglycerone phosphate synthase (AGPS) can decrease cancer malignancy, and the overexpression of them can the increase viability and migration potential of various tumor cell types; however, the role of AGPS in the proliferation and migration of glioma, and the association of Tudor-SN and AGPS in human glioma is not clear. In the present study, it was determined that AGPS silencing suppressed the proliferation and migration potential of glioma U87MG cells, and suppressed the expression of the circular RNAs circ-ubiquitin-associated protein 2, circ-zinc finger protein 292 and circ-homeodomain-interacting protein kinase 3, and the long non-coding RNAs H19 imprinted maternally expressed transcript (non-protein coding), colon cancer-associated transcript 1 (non-protein coding) and hepatocellular carcinoma upregulated long non-coding RNA. Furthermore, Tudor-SN silencing suppressed the expression of AGPS; however, nuclear factor (NF)-κB and microRNA (miR)-127 retrieval experiments partially reduced the expression of AGPS. Additionally, it was determined that Tudor-SN silencing suppressed the activity of the mechanistic target of rapamycin (mTOR) signaling pathway, and NF-κB and miR-127 retrieval experiments partially reduced the activity of mTOR. Therefore, it was considered that NF-κB and miR-127 may be the mediators of Tudor-SN-regulated AGPS via the mTOR signaling pathway. These results improve on our knowledge of the mechanisms underlying Tudor-SN and AGPS in human glioma.