Prior knowledge-guided multilevel graph neural network for tumor risk prediction and interpretation via multi-omics data integration.

The interrelation and complementary nature of multi-omics data can provide valuable insights into the intricate molecular mechanisms underlying diseases. However, challenges such as limited sample size, high data dimensionality and differences in omics modalities pose significant obstacles to fully harnessing the potential of these data. The prior knowledge such as gene regulatory network and pathway information harbors useful gene-gene interaction and gene functional module information. To effectively integrate multi-omics data and make full use of the prior knowledge, here, we propose a Multilevel-graph neural network (GNN): a hierarchically designed deep learning algorithm that sequentially leverages multi-omics data, gene regulatory networks and pathway information to extract features and enhance accuracy in predicting survival risk. Our method achieved better accuracy compared with existing methods. Furthermore, key factors nonlinearly associated with the tumor pathogenesis are prioritized by employing two interpretation algorithms (i.e. GNN-Explainer and IGscore) for neural networks, at gene and pathway level, respectively. The top genes and pathways exhibit strong associations with disease in survival analyses, many of which such as SEC61G and CYP27B1 are previously reported in the literature.

[Energy-saving and Emission Reduction Path for Road Traffic in Key Coastal Cities of Guangdong, Fujian and Zhejiang].

The rapid development of society and economy has resulted in a substantial increase in energy consumption, consequently exacerbating pollution issues. Current research predominantly focuses on energy-saving and emission reduction in road transportation within individual cities or the three major economic regions of China:the Yangtze River Delta, the Pearl River Delta, and the Beijing-Tianjin-Hebei Region. However, there is a dearth of studies addressing the southeastern coastal economic region. Located at the heart of China's southeastern coastal economic development, the provinces of Guangdong, Fujian, and Zhejiang unavoidably face challenges associated with energy consumption and emissions while pursuing economic growth. To address these challenges, this study employed a LEAP model to construct various scenarios for road transportation in the key coastal cities of Guangdong, Fujian, and Zhejiang from 2015 to 2035. These scenarios included a baseline scenario (BAU), an existing policy scenario (EPS), and an improved policy scenario (MPS). The MPS and EPS encompassed vehicle structure optimization (VSO), improved fuel economy (IFE), and reduced annual average mileage (RDM). By simulating and evaluating these scenarios, the energy-saving and emission reduction potentials of road transportation in the key coastal cities were assessed. The results indicated that, in the primary scenario, the MPS exhibited the most significant improvements in energy-saving, carbon reduction, and pollutant reduction effects. By 2035, the MPS achieved a remarkable 75% energy-saving rate compared to that in the baseline scenario, accompanied by reductions of 68%, 59%, 66%, 70%, and 64% in CO2, CO, NOx, PM2.5, and SO2 emissions, respectively. In the secondary scenario, the improved scenario of enhancing fuel economy achieved a notable 30% reduction in energy consumption. Additionally, the scenarios involving vehicle structure adjustment (yielding reductions of 36%, 30%, 36%, 26%, and 40%) and annual average mileage reduction (resulting in reductions of 37%, 37%, 36%, 37%, and 36%) demonstrated significant reductions in CO2, CO, NOx, PM2.5, and SO2 emissions.

Spatio-temporal evolution characteristics of carbon emissions from road transportation in the mainland of China from 2006 to 2021.

The leaping forward of the economy has promoted the rapid growth of road traffic demand, resulting in the carbon emissions of road traffic increasing significantly. It is well known that a one-size-fits-all emission reduction policy is not feasible. Therefore, conducting an investigation on the carbon emissions of all provincial-level regions within a country can assist the government in formulating carbon emission policies at a macro level tailored to different regions. In this study, the whole provincial-level administrative regions in the mainland of China were selected to quantify the carbon emissions of road traffic, and the carbon emissions from 2006 to 2021 were obtained by employing the top-down model. What's more, spatiotemporal characteristics of road transportation carbon emissions in those regions were explored based on Moran's I spatial autocorrelation method. In addition, the LMDI model was constructed based on five driving factors, namely energy intensity, energy consumption intensity, industrial scale, economic development, and population size, and the decomposition analysis of driving factors is carried out. The results show that carbon emissions from road traffic in all provincial regions showed an overall rising trend in the research period, with an average annual growth rate of 11.83 %. The distribution of road transportation carbon emissions exhibited an east-high, west-low distribution, with significantly higher emissions in the eastern and coastal regions compared to inland areas, additionally, China's seven geographical regions showed an initial rapid increase in carbon emissions followed by a stable growth trend. Secondly, five types of spatial clustering were identified of carbon emissions within provincial regions. Thirdly, the impacts of energy intensity and industrial scale were detrimental to road transportation carbon emissions, whereas economic development, energy consumption intensity, and population size had contrasting effects. Implications according to the above conclusions were put forward, aiming to provide guidance for the sustainable development of road transportation and expediting the achievement of the "carbon peaking and carbon neutrality" objective.

DERF: distinctive efficient robust features from the biological modeling of the P ganglion cells.

Studies in neuroscience and biological vision have shown that the human retina has strong computational power, and its information representation supports vision tasks on both ventral and dorsal pathways. In this paper, a new local image descriptor, termed distinctive efficient robust features (DERF), is derived by modeling the response and distribution properties of the parvocellular-projecting ganglion cells in the primate retina. DERF features exponential scale distribution, exponential grid structure, and circularly symmetric function difference of Gaussian (DoG) used as a convolution kernel, all of which are consistent with the characteristics of the ganglion cell array found in neurophysiology, anatomy, and biophysics. In addition, a new explanation for local descriptor design is presented from the perspective of wavelet tight frames. DoG is naturally a wavelet, and the structure of the grid points array in our descriptor is closely related to the spatial sampling of wavelets. The DoG wavelet itself forms a frame, and when we modulate the parameters of our descriptor to make the frame tighter, the performance of the DERF descriptor improves accordingly. This is verified by designing a tight frame DoG, which leads to much better performance. Extensive experiments conducted in the image matching task on the multiview stereo correspondence data set demonstrate that DERF outperforms state of the art methods for both hand-crafted and learned descriptors, while remaining robust and being much faster to compute.