Immunogenicity of a spike protein subunit-based COVID-19 vaccine with broad protection against various SARS-CoV-2 variants in animal studies.

SARS-CoV-2 pandemic has profound impacts on human life and global economy since the outbreak in 2019. With the new variants continue to emerge with greater immune escaping capability, the protectivity of the available vaccines is compromised. Therefore, development a vaccine that is capable of inducing immunity against variants including omicron strains is in urgent need. In this study, we developed a protein-based vaccine BCVax that is consisted of antigen delta strain spike protein and QS21-based adjuvant AB801 in nanoparticle immune stimulation complex format (AB801-ISCOM). Results from animal studies showed that high level of anti-S protein IgG was induced after two doses of BCVax and the IgG was capable of neutralizing multiple variants of pseudovirus including omicron BA.1 or BA.2 strains. In addition, strong Th1 response was stimulated after BCVax immunization. Furthermore, BCvax with AB801-ISCOM as the adjuvant showed significant stronger immunity compared with the vaccine using aluminum hydroxide plus CpG 1018 as the adjuvant. BCVax was also evaluated as a booster after two prior vaccinations, the IgG titers and pseudovirus neutralization activities against BA.2 or BA.4/BA.5 were further enhanced suggesting BCVax is a promising candidate as booster. Taken together, the pre-clinical data warrant BCVax for further development in clinic.

Model Pruning Enables Efficient Federated Learning on Edge Devices.

Federated learning (FL) allows model training from local data collected by edge/mobile devices while preserving data privacy, which has wide applicability to image and vision applications. A challenge is that client devices in FL usually have much more limited computation and communication resources compared to servers in a data center. To overcome this challenge, we propose PruneFL -a novel FL approach with adaptive and distributed parameter pruning, which adapts the model size during FL to reduce both communication and computation overhead and minimize the overall training time, while maintaining a similar accuracy as the original model. PruneFL includes initial pruning at a selected client and further pruning as part of the FL process. The model size is adapted during this process, which includes maximizing the approximate empirical risk reduction divided by the time of one FL round. Our experiments with various datasets on edge devices (e.g., Raspberry Pi) show that: 1) we significantly reduce the training time compared to conventional FL and various other pruning-based methods and 2) the pruned model with automatically determined size converges to an accuracy that is very similar to the original model, and it is also a lottery ticket of the original model.

Salivary gland tumors: A 20-year review of clinical diagnostic accuracy at a single center.

Obtaining reliable pre-operative diagnosis is crucial in planning treatment for patients with salivary gland tumors. The purpose of this study was to evaluate the accuracy of pre-operative clinical diagnosis of salivary gland tumors managed at a single tertiary university hospital over a period of 20 years. A retrospective analysis of the period between 1992 and 2011 was carried out to review the cases of patients with salivary gland tumors. A total of 101 patients were enrolled and general data were described. Clinical diagnosis was compared with the final pathological diagnosis to reveal the clinical diagnostic accuracy. Of the parotid and submandibular gland tumors, 86 and 67% were benign, respectively. The clinical diagnostic accuracies for diagnosis of parotid tumors as benign or malignant were 100 and 57%, respectively. The clinical diagnostic accuracies for diagnosis of submandibular tumors as benign or malignant were 67%. Therefore, the overall clinical judgment of benign and malignant tumors in the submandibular gland is unreliable. The accuracy for a parotid tumor to be clinically interpreted as benign was 100%. While it is difficult to draw any conclusion for non-parotid gland tumors, surgical intervention should be recommended in patients with parotid tumors clinically suspected to be malignant, and all submandibular, sublingual and minor salivary gland tumors.

Staggered multistep elution solid-phase extraction capillary electrophoresis/tandem mass spectrometry: a high-throughput approach in protein analysis.

An approach based on staggered multistep elution solid-phase extraction (SPE) capillary electrophoresis/tandem mass spectrometry (CE/MS/MS) was developed in the analysis of digested protein mixtures. On-line coupling of SPE with CE/MS was achieved using a two-leveled two-cross polydimethylsiloxane (PDMS)-based interface. Multistep elution SPE was used prior to CE to provide an additional dimension of separation, thus extending the separation capacity for the peptide mixture analysis. By decreasing in the number of co-eluting peptides, problems stemming from ionization suppression and finite MS/MS duty cycle were reduced. As a result, sequence coverage increased significantly using multistep elution SPE-CE/MS/MS compared to one-step elution SPE-CE/MS/MS in the analysis of a single protein tryptic digest (49% vs. 18%) and a six protein tryptic digest (22-71% vs. 10-44%). A staggered CE method was incorporated to increase the throughput. The electropherograms of consecutive CE runs were partially overlapped by injecting the sample plug at a fixed time interval. With the use of a 5 min injection interval, slightly poor results were obtained in comparison with the sequential CE method while the total analysis time was reduced to 28%.

The development of a two-leveled two cross interface for on-line coupling solid-phase extraction and capillary electrophoresis-mass spectrometry.

A two-leveled, two cross PDMS-based interface for on-line coupling of SPE with CE-MS was proposed. In this interface, the SPE column and the CE separation column were positioned orthogonally and two crosses were fabricated on the interface. With the two cross design, the operation of SPE could be performed independently without unexpected flow through leakage into the separation column. The performance of the interface was optimized using a peptide mixture. The position of the SPE column related to the CE separation channel was found to be critical to the performance of the system. Under the optimal position, the separation efficiency was similar to a CE-MS experiment without SPE. The peptide signals were enhanced 50- to 100-fold and the repeatability was within 4% RSD for migration time and 10% RSD for peak area. A tryptic digest of cytochrome c was used to demonstrate the feasibility of the interface in protein identification at a level of 1 ng/microL.