Spatial-temporal graph neural ODE networks for skeleton-based action recognition.

In the field of skeleton-based action recognition, accurately recognizing human actions is crucial for applications such as virtual reality and motion analysis. However, this task faces challenges such intraindividual action differences and long-term temporal dependencies. To address these challenges, we propose an innovative model called spatial-temporal graph neural ordinary differential equations (STG-NODE). First, in the data preprocessing stage, the dynamic time warping (DTW) algorithm is used to normalize and calculate 3D skeleton data to facilitate the derivation of customized adjacency matrices for improving the influence of intraindividual action differences. Secondly, a custom ordinary differential equation (ODE) integrator is applied based on the initial conditions of the temporal features, producing a solution function that simulates the dynamic evolution trend of the events of interest. Finally, the outstanding ODE solver is used to numerically solve the time features based on the solution function to increase the influence of long-term dependencies on the recognition accuracy of the model and provide it with a more powerful temporal modeling ability. Through extensive experiments conducted on the NTU RGB+D 60 and Kinetics Skeleton 400 benchmark datasets, we demonstrate the superior performance of STG-NODE in the action recognition domain. The success of the STG-NODE model also provides new ideas and methods for the future development of the action recognition field.

Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) moonlights as an adhesin in Mycoplasma hyorhinis adhesion to epithelial cells as well as a plasminogen receptor mediating extracellular matrix degradation.

Mycoplasma hyorhinis infects pigs causing polyserositis and polyarthritis, and has also been reported in a variety of human tumor tissues. The occurrence of disease is often linked with the systemic invasion of the pathogen. Glyceraldehyde-3-Phosphate Dehydrogenase (GAPDH), one of the key enzymes of glycolysis, was reported as a surface multifunctional molecule in several bacteria. Here, we investigated whether GAPDH could manifest binary functions; as an adhesin to promote colonization as well as a plasminogen receptor functioning in extracellular matrix (ECM) degradation to promote systemic invasion. The surface localization of GAPDH was observed in M. hyorhinis with flow cytometry and colony blot analysis. Recombinant GAPDH (rGAPDH) was found to be able to bind porcine-derived PK-15 and human-derived NCI-H292 cells. The incubation with anti-GAPDH antibody significantly decreased the adherence of M. hyorhinis to both cell lines. To investigate its function in recruiting plasminogen, firstly, the interaction between rGAPDH and plasminogen was demonstrated by ELISA and Far-Western blot assay. The activation of the rGAPDH-bound plasminogen into plasmin was proved by using a chromogenic substrate, and furtherly confirmed to degrade extracellular matrix by using a reconstituted ECM. Finally, the ability of rGAPDH to bind different ECM components was demonstrated, including fibronectin, laminin, collagen type IV and vitronectin. Collectively, our data imply GAPDH as an important adhesion factor of M. hyrohinis and a receptor for hijacking host plasminogen to degrade ECM. The multifunction of GAPDH to bind both plasminogen and ECM components is believed to increase the targeting of proteolysis and facilitate the dissemination of M. hyorhinis.