Abstract Integrating single-cell datasets produced by multiple omics technologies is essential for defining cellular heterogeneity. Mosaic integration, in which different share only some of the measured modalities, poses major challenges, particularly regarding modality alignment and batch effect removal. Here, we present a deep probabilistic framework mosaic integration knowledge transfer (MIDAS) multimodal data. MIDAS simultaneously achieves dimensionality reduction, imputation correction...

Abstract Motivation Retrosynthesis is a critical task in drug discovery, aimed at finding viable pathway for synthesizing given target molecule. Many existing approaches frame this as graph-generating problem. Specifically, these methods first identify the reaction center, and break targeted molecule accordingly to generate synthons. Reactants are generated by either adding atoms sequentially synthon graphs or directly appropriate leaving groups. However, both of strategies have limitations....

Predicting drug-target binding affinity is critical for drug discovery, as it helps identify promising candidates and predict their effectiveness. Recent advancements in deep learning have made significant progress tackling this task. However, existing methods heavily rely on training data, performance often limited when predicting affinities new drugs targets. To address challenge, we propose a novel Generalized Feature Learning (GFLearn) model prediction. By integrating Graph Neural...

Proteins dynamically interact with each other to perform their biological functions. The dynamic operations of protein interaction networks (PPI) are also reflected in the formations complexes. Existing complex detection algorithms usually overlook inherent temporal nature interactions within PPI networks. Systematically analyzing complexes can not only improve accuracy detection, but strengthen our knowledge on assembly processes for cellular organization.In this study, we propose a novel...

Disease-gene association through Genome-wide study (GWAS) is an arduous task for researchers. Investigating single nucleotide polymorphisms (SNPs) that correlate with specific diseases needs statistical analysis of associations. Considering the huge number possible mutations, in addition to its high cost, another important drawback GWAS large false-positives. Thus, researchers search more evidence cross-check their results different sources. To provide alternative low-cost disease-gene...

Abstract Motivation Predicting potential links in biomedical bipartite networks can provide useful insights into the diagnosis and treatment of complex diseases discovery novel drug targets. Computational methods have been proposed recently to predict for various networks. However, existing are usually rely on coverage known links, which may encounter difficulties when dealing with new nodes without any link information. Results In this study, we propose a prediction method, named graph...

Mutation-induced variation of protein-ligand binding affinity is the key to many genetic diseases and emergence drug resistance, therefore predicting such mutation impacts great importance. In this work, we aim predict on using efficient structure-based, computational methods.Relying consolidated databases experimentally determined data characterize change upon based a number local geometrical features monitor feature differences during molecular dynamics (MD) simulations. The are quantified...

Recently, several studies have drawn attention to the determination of a minimum set driver proteins that are important for control underlying protein-protein interaction (PPI) networks. In general, dominating (MDS) model is widely adopted. However, because MDS does not generate unique configuration, multiple different MDSs would be generated when using optimization algorithms. Therefore, among these MDSs, it difficult find out one represents true proteins.To address this problem, we develop...

Abstract Motivation Single-cell clustering plays a crucial role in distinguishing between cell types, facilitating the analysis of heterogeneity mechanisms. While many existing methods rely solely on gene expression data obtained from single-cell RNA sequencing techniques to identify clusters, information contained mono-omic is often limited, leading suboptimal performance. The emergence multi-omics technologies enables integration multiple omics for identifying but how integrate different...

Understanding how gene regulatory networks change under different cellular states is important for revealing insights into network dynamics. Gaussian graphical models, which assume that the data follow a joint normal distribution, have been used recently to infer differential networks. However, distributions of omics are non-normal in general. Furthermore, although much biological knowledge (or prior information) has accumulated, most existing methods ignore valuable information. Therefore,...

Identification of protein complexes is fundamental for understanding the cellular functional organization. With accumulation physical protein-protein interaction (PPI) data, computational detection from available PPI networks has drawn a lot attentions. While most existing complex algorithms focus on analyzing network, none them take into account "signs" (i.e., activation-inhibition relationships) interactions. As interactions reflect way proteins communicate, considering can not only...

Abstract Background Synthetic lethality has attracted a lot of attentions in cancer therapeutics due to its utility identifying new anticancer drug targets. Identifying synthetic lethal (SL) interactions is the key step towards exploration treatment. However, biological experiments are faced with many challenges when interactions. Thus, it necessary develop computational methods which could serve as useful complements experiments. Results In this paper, we propose novel graph regularized...

Understanding how the structure of gene dependency network changes between two patient-specific groups is an important task for genomic research. Although many computational approaches have been proposed to undertake this task, most them estimate correlation networks from group-specific expression data independently without considering common shared different groups. In addition, with development high-throughput technologies, we can collect profiles same patients multiple platforms....

Abstract Summary Imputation of dropout events that may mislead downstream analyses is a key step in analyzing single-cell RNA-sequencing (scRNA-seq) data. We develop EnImpute, an R package introduces ensemble learning method for imputing scRNA-seq EnImpute combines the results obtained from multiple imputation methods to generate more accurate result. A Shiny application developed provide easier implementation and visualization. Experiment show outperforms individual state-of-the-art almost...

Many Gram-negative bacteria infect hosts and cause diseases by translocating a variety of type III secreted effectors (T3SEs) into the host cell cytoplasm. However, despite dramatic increase in number available whole-genome sequences, it remains challenging for accurate prediction T3SEs. Traditional models have focused on atypical sequence features buried N-terminal peptides T3SEs, but unfortunately, these had high false-positive rates. In this research, we integrated promoter information...

Gene regulatory networks (GRNs) unveil the intricate interactions among genes, pivotal in elucidating complex biological processes within cells. The advent of single-cell RNA-sequencing (scRNA-seq) enables inference GRNs at resolution. However, majority current supervised network methods typically concentrate on predicting pairwise gene interaction, thus failing to fully exploit correlations all genes and exhibiting limited generalization performance. To address these issues, we propose a...

Contrastive clustering performs and data representation in a unified model, where instance- cluster-level constrastive learning are conducted simultaneously. However, commonly-used augmentation methods make contrastive mechanism effect but may cause getting stuck domain-specific information, which further deteriorates performance limits generalization ability. To this end, we propose new framework, named Generalized Clustering with domain shifts modeling (GeCC), can integrate diverse...

Abstract Motivation The identification of repetitive elements is important in genome assembly and phylogenetic analyses. existing de novo repeat methods exploiting the use short reads are impotent identifying long repeats. Since more likely to cover regions completely, using favorable for recognizing Results In this study, we propose a novel method namely RepLong based on PacBio reads. Given that mapped highly overlapped with each other, equivalent discovery consensus overlaps between reads,...

Predicting disease causative genes (or simply, genes) has played critical roles in understanding the genetic basis of human diseases and further providing treatment guidelines. While various computational methods have been proposed for gene prediction, with recent increasing availability biological information genes, it is highly motivated to leverage these valuable data sources extract useful accurately predicting genes. We present an integrative framework called N2VKO predict Firstly, we...

Exploring the rewiring pattern of gene regulatory networks between different pathological states is an important task in bioinformatics. Although a number computational approaches have been developed to infer differential from high-throughput data, most them only focus on expression data. The valuable static network data accumulated recent biomedical researches are neglected. In this study, we propose new Gaussian graphical model-based method by integrating and We first evaluate empirical...

Single-cell RNA sequencing (scRNA-seq) methods make it possible to reveal gene expression patterns at single-cell resolution. Due technical defects, dropout events in scRNA-seq will add noise the gene-cell matrix and hinder downstream analysis. Therefore, is important for recovering true levels before carrying out analysis.In this article, we develop an imputation method, called scTSSR, recover scRNA-seq. Unlike most existing that impute by borrowing information across only genes or cells,...

Mild cognitive impairment (MCI) is an early stage of Alzheimer's disease (AD), which a neurodegenerative disease. Functional connectivity networks (FCN) provide effective method for analyzing brain functional regions connectivity. However, most methods only considered the neuroimaging information and focused on group relationship without subjects' individual features, ignored demographic relationship. To handle it, in this paper, we introduce novel based graph convolutional (GCN), combines...

This study aims to investigate healthcare workers' (HCWs) willingness receive SARS-CoV-2 vaccine in Zhejiang and discover the related influential factors. The survey was conducted six regions of Province, China, 13 hospitals 12 Centers for Disease Control Prevention (CDC) were incorporated into research. Participants workers a total 3726 questionnaires collected online, which 3634 (97.53%) analyzed. relationships between factors get vaccinated against COVID-19 computed as odds ratios (ORs)...