Search

The Hi-C technique has enabled genome-wide mapping of chromatin interactions and investigated the organizational principles of the spatial structure of the genome. However, computational methods for studying Hi-C are still in the early stage.Pioneering models have been developed to reconstruct the 3D genome structures. But there is no consistent measuring result between the modeling methods, thereby imposing challenges for the users to interpret the organization of the 3D structures and to understand the genome functions. Moreover, 3D genome modeling becomes more complicated at higher resolution because of the sparsity and diversity in bulk Hi-C and the restrictions of computational resources. Furthermore, high-resolution Hi-C requires costly, deep sequencing; therefore, it has only been achieved for a limited number of cell types. Neural networks have been developed as a remedy to these problems at high resolution.In this work, we first reviewed the 3D structure modeling methods comprehensively. We developed a simulation method based on the 3D conformations from single-cell modeling and several evaluation metrics for measuring the similarity between structures. We profiled the performance of existing bulk Hi-C based 3D genome structure modeling methods using both simulated and real bulk Hi-C.Next, we proposed a novel method, GIST, for predicting 3D structures at a high resolution based on Auto-encoder with GAT. We convert the Hi-C into a heterogeneous graph, and GIST encodes the graph as a population of 3D conformations optimized by edge classification. We demonstrated that GIST produced chromosome structures consistent with FISH and outperformed existing 3D modeling methods. We illustrated the diversity of 3D structure predictions by evaluating the active and inactive X chromosome structures.Lastly, we proposed a novel method, EnHiC, for predicting high-resolution Hi-C from low-resolution input based on a generative adversarial network. Inspired by non-negative matrix factorization, EnHiC fully exploits the unique properties of Hi-C and extracts rank-1 features from multi-scale low-resolution matrices to enhance the resolution. We demonstrated that EnHiC accurately and reliably enhanced the resolution of Hi-C and outperformed other GAN-based models. EnHiC-predicted high-resolution matrices facilitated the accurate detection of topologically associated domains and fine-scale chromatin interactions.

To understand website complexity deeply, a web page complexity measurement system is developed. The system measures the complexity of a web page at two levels: transport-level and content-level, using a packet trace-based approach rather than server or client logs. Packet traces surpass others in the amount of information contained. Quantitative analyses show that different categories of web pages have different complexity characteristics. Experimental results show that a news web page usually loads much more elements at more accessing levels from much more web servers within diverse administrative domains over much more concurrent transmission control protocol (TCP) flows. About more than half of education pages each only involve a few logical servers, where most of elements of a web page are fetched only from one or two logical servers. The number of content types for web game traffic after login is usually least. The system can help web page designers to design more efficient web pages, and help researchers or Internet users to know communication details.

In the process of image segmentation based on clustering, it is necessary to manually determine the number of clusters segmented, which makes the image automation less. A fuzzy clustering algorithm is proposed to automatically determine the number of clusters. First, the similarity matrix is constructed using the features of the extracted image, and the number of clusters is determined using the AP clustering algorithm. Then, the obtained number of clusters is input to the FCM algorithm, and the clustering center is optimized using the firefly algorithm. Finally, a simulation experiment was conducted. The experimental results show that the improved FCM algorithm has a good effect and can realize automatic image segmentation.

Objective: To design the lip-color features and classification model and provide an automatic, quantitative method based on the lip detection in facial image. Methods: In this paper, We adopted the lip segmentation algorithm based on the three-dimensional mixture Skin Gaussian Model and color classification in SVM to solve this problem. Specifically, we used the GMM based iterations to confirm skin-color pixels that belong to the skin area. In the process of optimization, we extracted two-dimensional GMM with five parameters as the characteristic to obtain the main color and secondary color to extract and classify the features of the 877 lip images. Consequence: The accuracy and stability of classification method was successfully proved. Conclusion: Firstly, We proposed a Skin Gaussian mixture Model that can be used for the robust and accurate segmentation of facial skin in a probabilistic manner. Secondly, we extracted five parameters from the two-dimensional GMM of the a, b component in the laboratory space as the characteristic of each picture. Finally, we classified the lips through the SVM classifier. As for the lips with mixed color, the method we proposed can have a good performance in identifying the secondary colors, thereby improving the identification precision of the image segmentation and color classification of the lip.

The health management service mode in traditional medicine for the public has attracted more and more attentions in recent years. With the development of large number of health related apps and devices, large quantities of valuable health data can be transmitted and collected through electronic devices such as smart phones. Accordingly, the quantization and digitization of the information in the four diagnostic methods in traditional Chinese medicine have played a more important role at present. In which the inspection diagnosis technology in traditional Chinese medicine can conduct effective diagnosis on health conditions by combining with modern technologies at the premise of reserving the features of traditional medicine. Therefore, the quantization of inspection diagnosis information is especially important. This article mainly introduces the method of tongue diagnosis information quantization in quantization of inspection diagnosis information.

Clustering trajectory data is an important way to mine hidden information behind moving object sampling data, such as understanding trends in movement patterns, gaining high popularity in geographic information and so on. In the era of `Big data', the current approaches for clustering trajectory data generally do not apply for excessive costs in both scalability and computing performance for trajectory big data. Aiming at these problems, this study first proposes a new clustering algorithm for trajectory big data, namely Tra-POPTICS by modifying a scalable clustering algorithm for point data (POPTICS). Tra-POPTICS has employed the spatiotemporal distance function and trajectory indexing to support trajectory data. Tra-POPTICS can process the trajectory big data in a distributed manner to meet a great scalability. Towards providing a fast solution to clustering trajectory big data, this study has explored the feasibility to utilize the contemporary general-purpose computing on the graphics processing unit (GPGPU). The GPGPU-aided clustering approach parallelized the Tra-POPTICS with the Hyper-Q feature of Kelper GPU and massive GPU threads. The experimental results indicate that (1) the Tra-POPTICS algorithm has a comparable clustering quality with T-OPTICS (the state of art work of clustering trajectories in a centralized fashion) and outperforms T-OPTICS by average four times in terms of scalability, and (2) the G-Tra-POPTICS has a comparable clustering quality with T-POPTICS as well and further gains about 30 speedup on average for clustering trajectories comparing to Tra-POPTICS with eight threads. The proposed algorithms exhibit great scalability and computing performance in clustering trajectory big data.