A5086316879- Complex Network Analysis Techniques
- Data Management and Algorithms
- Bioinformatics and Genomic Networks
- Computational Drug Discovery Methods
- Opinion Dynamics and Social Influence
- Peer-to-Peer Network Technologies
- Advanced Database Systems and Queries
- Machine Learning in Bioinformatics
- Protein Structure and Dynamics
- Service-Oriented Architecture and Web Services
- Domain Adaptation and Few-Shot Learning
- Gene expression and cancer classification
- Theoretical and Computational Physics
- Bayesian Modeling and Causal Inference
- RNA and protein synthesis mechanisms
- Genomics and Chromatin Dynamics
- Advanced Neural Network Applications
- Anomaly Detection Techniques and Applications
- Semantic Web and Ontologies
- Caching and Content Delivery
- Text and Document Classification Technologies
- Topic Modeling
- Geographic Information Systems Studies
- Network Security and Intrusion Detection
- Mobile Agent-Based Network Management
Tongji University
2016-2025
Fudan University
2009-2023
Zhejiang Lab
2023
ORCID
2021
Robert Bosch (Germany)
2021
Chinese PLA General Hospital
2018
Changzhou University
2013-2015
Nanyang Technological University
2015
Howard Hughes Medical Institute
2011
Central South University
2007
Abstract Motivation: Computational prediction of compound–protein interactions (CPIs) is great importance for drug design and development, as genome-scale experimental validation CPIs not only time-consuming but also prohibitively expensive. With the availability an increasing number validated interactions, performance computational approaches severely impended by lack reliable negative CPI samples. A systematic method screening sample becomes critical to improving in silico methods....
Abstract Single-cell transcriptomic assays have enabled the de novo reconstruction of lineage differentiation trajectories, along with characterization cellular heterogeneity and state transitions. Several methods been developed for reconstructing developmental trajectories from single-cell data, but efforts on analyzing epigenomic data trajectory visualization remain limited. Here we present STREAM, an interactive pipeline capable disentangling visualizing complex branching both data. We...
Recently, "pre-training and fine-tuning'' has been adopted as a standard workflow for many graph tasks since it can take general knowledge to relieve the lack of annotations from each application. However, with node level, edge level are far diversified, making pre-training pretext often incompatible these multiple tasks. This gap may even cause "negative transfer'' specific application, leading poor results. Inspired by prompt learning in natural language processing (NLP), which presented...
Abstract Motivation: Fold recognition is an important step in protein structure and function prediction. Traditional sequence comparison methods fail to identify reliable homologies with low identity, while the taxonomic are effective alternatives, but their prediction accuracies around 70%, which still relatively for practical usage. Results: In this study, a simple powerful method presented fold recognition, combines support vector machine (SVM) autocross-covariance (ACC) transformation....
Accurately predicting drug-target interaction (DTI) is a crucial step to drug discovery. Recently, deep learning techniques have been widely used for DTI prediction and achieved significant performance improvement. One challenge in building models how appropriately represent drugs targets. Target distance map molecular graph are low dimensional informative representations, which however not jointly prediction. Another effectively model the mutual impact between Though attention mechanism has...
Single-cell RNA sequencing (scRNA-seq) permits researchers to study the complex mechanisms of cell heterogeneity and diversity. Unsupervised clustering is central importance for analysis scRNA-seq data, as it can be used identify putative types. However, due noise impacts, high dimensionality pervasive dropout events, data remains a computational challenge. Here, we propose new deep structural method named scDSC, which integrate information into single cells. The proposed scDSC consists...
The generic object detection (GOD) task has been successfully tackled by recent deep neural networks, trained an avalanche of annotated training samples from some common classes. However, it is still non-trivial to generalize these detectors the novel long-tailed classes, which have only few labeled samples. To this end, Few-Shot Object Detection (FSOD) topical recently, as mimics humans’ ability learning learn and intelligently transfers learned knowledge heavy-tailed Especially, research...
The explicit determinations of the mean first-passage time (MFPT) for trapping problem are limited to some simple structure, e.g., regular lattices and geometrical fractals, determining MFPT random walks on other media, especially complex real networks, is a theoretical challenge. In this paper, we investigate walk pseudofractal scale-free web (PSFW) with perfect trap located at node highest degree, which simultaneously exhibits remarkable small-world properties observed in networks. We...
We describe PredUs, an interactive web server for the prediction of protein–protein interfaces. Potential interfacial residues a query protein are identified by 'mapping' contacts from known interfaces protein's structural neighbors to surface query. calculate score each residue be with support vector machine. Results can visualized in molecular viewer and number features allow users tailor particular hypothesis. The PredUs is available at:...
MicroRNAs (simply miRNAs) are derived from larger hairpin RNA precursors and play essential regular roles in both animals plants. A number of computational methods for miRNA genes finding have been proposed the past decade, yet problem is far being tackled, especially when considering imbalance issue known miRNAs unidentified miRNAs, pre-miRNAs with multi-loops or higher minimum free energy (MFE). This paper presents a new approach, miRenSVM, genes. Aiming at better prediction performance,...
A vast variety of real-life networks display the ubiquitous presence scale-free phenomenon and small-world effect, both which play a significant role in dynamical processes running on networks. Although various have been investigated networks, analytical research about random walks such is much less. In this paper, we will study analytically scaling mean first-passage time (MFPT) for To end, first map classical Koch fractal to network, called network. According proposed mapping, present an...
Since traditional drug research and development is often time-consuming high-risk, there an increasing interest in establishing new medical indications for approved drugs, referred to as repositioning, which provides a relatively low-cost high-efficiency approach discovery. With the explosive growth of large-scale biochemical phenotypic data, repositioning holds great potential precision medicine post-genomic era. It urgent develop rational systematic approaches predict drugs on large...
Identifying specific hot spot residues that contribute significantly to the affinity and specificity of protein interactions is a problem utmost importance. We present an interactive web server, PredHS, which based on effective structure-based prediction method. The PredHS method integrates many novel structural energetic features with two types neighborhoods (Euclidian Voronoi), combines random forest sequential backward elimination algorithms select optimal subset features. achieved...
Molecular property prediction is a hot topic in recent years. Existing graph-based models ignore the hierarchical structures of molecules. According to knowledge chemistry and pharmacy, functional groups molecules are closely related its physio-chemical properties binding affinities. So, it should be helpful represent molecular graphs by fragments that contain for prediction.In this article, boost performance molecule prediction, we first propose definition graph may or groups, which...
Most object detection methods require huge amounts of annotated data and can detect only the categories that appear in training set. However, reality acquiring massive is both expensive time-consuming. In this paper, we propose a novel two-stage detector for accurate few-shot detection. first stage, employ support-query mutual guidance mechanism to generate more support-relevant proposals. Concretely, on one hand, query-guided support weighting module developed aggregating different supports...
Many real networks share three generic properties: they are scale-free, display a small-world effect, and show power-law strength-degree correlation. In this paper, we propose type of deterministically growing called Sierpinski networks, which induced by the famous fractals constructed in simple iterative way. We derive analytical expressions for degree distribution, strength clustering coefficient, correlation, agree well with characterizations various real-life networks. Moreover, that...
The family of Vicsek fractals is one the most important and frequently studied regular fractal classes, it considerable interest to understand dynamical processes on this treelike family. In paper, we investigate discrete random walks fractals, with aim obtain exact solutions global mean-first-passage time (GMFPT), defined as average first-passage (FPT) between two nodes over whole fractals. Based known connections FPTs, effective resistance, eigenvalues graph Laplacian, determine implicitly...