To better understand the biological significance of Ca(2+), we report a comprehensive statistical analysis of calcium-binding proteins from the Protein Data Bank to identify structural parameters associated with EF-hand and non-EF-hand Ca(2+)-binding sites. Comparatively, non-EF-hand sites utilize lower coordination numbers (6 +/- 2 vs. 7 +/- 1), fewer protein ligands (4 +/- 2 vs. 6 +/- 1), and more water ligands (2 +/- 2 vs. 1 +/- 0) than EF-hand sites. The orders of ligand preference for non-EF-hand and EF-hand sites, respectively, were H(2)O (33.1%) > side-chain Asp (24.5%) > main-chain carbonyl (23.9%) > side-chain Glu (10.4%), and side-chain Asp (29.7%) > side-chain Glu (26.6%) > main-chain carbonyl (21.4%) > H(2)O (13.3%). Less formal negative charge was observed in the non-EF-hand than in the EF-hand binding sites (1 +/- 1 vs. 3 +/- 1). Additionally, over 20% of non-EF-hand sites had formal charge values of zero due to increased utilization of water and carbonyl oxygen ligands. Moreover, the EF-hand sites presented a narrower range of ligand distances and bond angles than non-EF-hand sites, possibly owing to the highly conserved helix-loop-helix motif. Significant differences between ligand types (carbonyl, side chain, bidentate) demonstrated that angles associated with each type must be classified separately, and the EF-hand side-chain Ca-O-C angles exhibited an unusual bimodal quality consistent with an Asp distribution that differed from the Gaussian model observed for non-EF-hand proteins. The results of this survey more accurately describe differences between EF-hand and non-EF-hand proteins and provide new parameters for the prediction and design of different classes of Ca(2+)-binding proteins.

Identifying calcium-binding sites in proteins is one of the first steps towards predicting and understanding the role of calcium in biological systems for protein structure and function studies. Due to the complexity and irregularity of calcium-binding sites, a fast and accurate method for predicting and identifying calcium-binding protein is needed. Here we report our development of a new fast algorithm (GG) to detect calcium-binding sites. The GG algorithm uses a graph theory algorithm to find oxygen clusters of the protein and a geometric algorithm to identify the center of these clusters. A cluster of four or more oxygen atoms has a high potential for calcium binding. High performance with about 90% site sensitivity and 80% site selectivity has been obtained for three datasets containing a total of 123 proteins. The results suggest that a sphere of a certain size with four or more oxygen atoms on the surface and without other atoms inside is necessary and sufficient for quickly identifying the majority of the calcium-binding sites with high accuracy. Our finding opens a new avenue to visualize and analyze calcium-binding sites in proteins facilitating the prediction of functions from structural genomic information.

The convolutional neural network (CNN) identification method and the BP neural network identification method were used to diagnose the bearing fault respectively. When using the CNN diagnostic method, first perform continuous wavelet transform (CWT) on the vibration signal of the rolling bearing to obtain a time-frequency map, and then compress the time-frequency map to an appropriate size; Then, the compressed time-frequency map is used as a feature map to input into the CNN classifier model established; finally, an experimental study is carried out based on the artificial bearing fault data set of Western Reserve University. The results show that the average accuracy rate of this method is greater than 99%.In the BP fault diagnosis method, based on the data set, nine parameters of average value, maximum value, minimum value, peak-to-peak value, root mean square value, standard deviation, variance, skewness and kurtosis are established as training input vectors. Using BP neural network with 10 hidden layer nodes for fault identification, the results show that the average accuracy of identification is 93.7839%. The comparative analysis of the two methods shows that the BP identification method has higher training efficiency and takes less time; the CNN identification method has higher recognition accuracy, but the training takes more time.

In recent years, high-resolution remote sensing semantic segmentation based on data fusion has gradually become a research focus in the field of land classification, which is an indispensable task of a smart city. However, the existing feature fusion methods with bottom-up structures can achieve limited fusion results. Alternatively, various auxiliary fusion modules significantly increase the complexity of the models and make the training process intolerably expensive. In this paper, we propose a new lightweight model called top-down pyramid fusion network (TdPFNet) including a multi-source feature extractor, a top-down pyramid fusion module and a decoder. It can deeply fuse features from different sources in a top-down structure using high-level semantic knowledge guiding the fusion of low-level texture information. Digital surface model (DSM) data and open street map (OSM) data are used as auxiliary inputs to the Potsdam dataset for the proposed model evaluation. Experimental results show that the network proposed in this paper not only notably improves the segmentation accuracy, but also reduces the complexity of the multi-source semantic segmentation model.