Essential proteins as a vital part of maintaining the cells' life play an important role in study biology and drug design. With generation large amounts biological data related to essential proteins, increasing number computational methods have been proposed. Different from which adopt single machine learning method or ensemble method, this paper proposes predicting framework named by XGBFEMF for identifying includes SUB-EXPAND-SHRINK constructing composite features with original obtaining...

Cancer is thought to be caused by the accumulation of driver genetic mutations. Therefore, identifying cancer genes plays a crucial role in understanding molecular mechanism and developing precision therapies biomarkers. In this work, we propose Multi-Task learning method, called MTGCN, based on Graph Convolutional Network identify genes. First, augment gene features introducing their protein-protein interaction (PPI) network. After that, multi-task framework propagates aggregates nodes...

Different cancer patients may respond differently to treatment due the heterogeneity of cancer. It is an urgent task develop efficient computational method identify drug responses in different cell lines, which guides us design personalized therapy for individual patient. Hence, we propose end-to-end algorithm, namely MOFGCN, predict response lines based on Multi-Omics Fusion and Graph Convolution Network. MOFGCN first fuses multiple omics data calculate line similarity then constructs a...

Essential proteins play an essential role in cell survival and replication. Currently, more computational methods are developed to identify proteins, which overcome the time-consuming, costly inefficient shortcomings with biological experimental methods.In order improve recognition rate, some new by fusing multiple features developed, but they seldom consider connection among features. After analyzing a large number of based on multi-feature fusion, phenomenon is found, called weak...

Prediction of essential proteins which are crucial to an organism's survival is important for disease analysis and drug design, as well the understanding cellular life. The majority prediction methods infer possibility be by using network topology. However, these limited completeness available protein-protein interaction (PPI) data depend on accuracy. To overcome limitations, some computational have been proposed. seldom them solve this problem taking consideration protein domains. In work,...

With the continuing development and improvement of genome-wide techniques, a great number candidate genes are discovered. How to identify most likely disease among large candidates becomes fundamental challenge in human health. A common view is that related specific or similar tend reside same neighbourhood biomolecular networks. Recently, based on such observations, many methods have been developed tackle this challenge. In review, we firstly introduce concept genes, their properties,...

Essential proteins are indispensable for cell survive. Identifying essential is very important improving our understanding the way of a working. There various types features related to essentiality proteins. Many methods have been proposed combine some them predict However, it still big challenge designing an effective method by integrating different features, and explaining how these selected decide protein. Gene expression programming (GEP) learning algorithm what learns specifically about...

Abstract Background Some proposed methods for identifying essential proteins have better results by using biological information. Gene expression data is generally used to identify proteins. However, gene prone fluctuations, which may affect the accuracy of protein identification. Therefore, we propose an identification method based on and PPI network calculate similarity "active" "inactive" state in a cluster network. Our experiments show that can improve predicting Results In this paper,...

Abstract Motivation Due to cancer heterogeneity, the therapeutic effect may not be same when a cohort of patients type receive treatment. The anticancer drug response prediction help develop personalized therapy regimens increase survival and reduce patients’ expenses. Recently, graph neural network-based methods have aroused widespread interest achieved impressive results on task. However, most them apply convolution process cell line-drug bipartite graphs while ignoring intrinsic...

Abstract The correct prediction of disease-associated miRNAs plays an essential role in disease prevention and treatment. Current computational methods to predict construct different miRNA views based on various properties then integrate the multiviews relationship between diseases. However, most existing ignore information interaction among consistency features (disease features) across multiple views. This study proposes a method hypergraph contrastive learning (MHCLMDA) miRNA–disease...

Identifying driver genes contributing to the occurrence and development of cancers plays a critical role in cancer research treatment. Some recent computational approaches identify cancer-driver based on gene networks, assuming that perform essential functions networks. Due noise function many works focus integrating networks derived from multi-omics datasets improve accuracy detection. However, most them ignore information interactions between these datasets. In this work, we propose MNGCL,...

Abstract Background Characterization of unknown proteins through computational approaches is one the most challenging problems in silico biology, which has attracted world-wide interests and great efforts. There have been some methods proposed to address this problem, are either based on homology mapping or context protein interaction networks. Results In paper, two algorithms by integrating protein-protein (PPI) network, proteins’ domain information complexes. The combination similarity...

Protein complexes play a significant role in understanding the underlying mechanism of most cellular functions. Recently, many researchers have explored computational methods to identify protein from protein-protein interaction (PPI) networks. One group focus on detecting local dense subgraphs which correspond by considering neighbors. The drawback this kind approach is that global information networks ignored. Some such as Markov Clustering algorithm (MCL), PageRank-Nibble are proposed find...

With the gap between sequence data and their functional annotations becomes increasing wider, many computational methods have been proposed to annotate functions for unknown proteins. However, designing effective make good use of various biological resources is still a big challenge researchers due function diversity In this work, we propose new method named ThrRW, which takes several steps random walking on three different networks: protein interaction network (PIN), domain co-occurrence...

MiRNAs are reported to be linked the pathogenesis of human complex diseases. Disease-related miRNAs may serve as novel bio-marks and drug targets. This work focuses on designing a multi-relational Graph Convolutional Network model predict miRNA-disease associations (HGCNMDA) from Heterogeneous network. HGCNMDA introduces gene layer construct miRNA-gene-disease heterogeneous We refine features nodes into initial inductive so that direct indirect between diseases miRNA can considered...

Conventional methods address the cross-modal retrieval problem by projecting multi-modal data into a shared representation space. Such strategy will inevitably lose modality-specific information, leading to decreased accuracy. In this paper, we propose heterogeneous graph embeddings preserve more abundant information. The embedding from one modality be compensated with aggregated other modality. particular, self-denoising tree search is designed reduce "label noise" problem, making...

Chest radiology imaging plays a crucial role in the early screening, diagnosis, and treatment of chest diseases. The accurate interpretation radiological images automatic generation reports not only save doctor's time but also mitigate risk errors diagnosis. core objective report is to achieve precise mapping visual features lesion descriptions at multi-scale fine-grained levels. Existing methods typically combine global textual generate reports. However, these approaches may ignore key...