Briefings in bioinformatics最新文献

Natural products (NPs) are indispensable in drug development, particularly in combating infections, cancer, and neurodegenerative diseases. However, their limited availability poses significant challenges. Template-free de novo biosynthetic pathway design provides a strategic solution for NP production, with deep learning standing out as a powerful tool in this domain. This review delves into state-of-the-art deep learning algorithms in NP biosynthesis pathway design. It provides an in-depth discussion of databases like Kyoto Encyclopedia of Genes and Genomes (KEGG), Reactome, and UniProt, which are essential for model training, along with chemical databases such as Reaxys, SciFinder, and PubChem for transfer learning to expand models' understanding of the broader chemical space. It evaluates the potential and challenges of sequence-to-sequence and graph-to-graph translation models for accurate single-step prediction. Additionally, it discusses search algorithms for multistep prediction and deep learning algorithms for predicting enzyme function. The review also highlights the pivotal role of deep learning in improving catalytic efficiency through enzyme engineering, which is essential for enhancing NP production. Moreover, it examines the application of large language models in pathway design, enzyme discovery, and enzyme engineering. Finally, it addresses the challenges and prospects associated with template-free approaches, offering insights into potential advancements in NP biosynthesis pathway design.

Spatial transcriptomics offers deep insights into cellular functional localization and communication by mapping gene expression to spatial locations. Traditional approaches that focus on selecting spatially variable genes often overlook the complexity of biological pathways and the interactions among genes. Here, we introduce a novel framework that shifts the focus towards directly identifying functional pathways associated with spatial variability by adapting the Brownian distance covariance test in an innovative manner to explore the heterogeneity of biological functions over space. Unlike most other methods, this statistical testing approach is free of gene selection and parameter selection and allows nonlinear and complex dependencies. It allows for a deeper understanding of how cells coordinate their activities across different spatial domains through biological pathways. By analyzing real human and mouse datasets, the method found significant pathways that were associated with spatial variation, as well as different pathway patterns among inner- and edge-cancer regions. This innovative framework offers a new perspective on analyzing spatial transcriptomic data, contributing to our understanding of tissue architecture and disease pathology. The implementation is publicly available at https://github.com/tianlq-prog/STpathway.

Substructure-based representation learning has emerged as a powerful approach to featurize complex attributed graphs, with promising results in molecular property prediction (MPP). However, existing MPP methods mainly rely on manually defined rules to extract substructures. It remains an open challenge to adaptively identify meaningful substructures from numerous molecular graphs to accommodate MPP tasks. To this end, this paper proposes Prototype-based cOntrastive Substructure IdentificaTion (POSIT), a self-supervised framework to autonomously discover substructural prototypes across graphs so as to guide end-to-end molecular fragmentation. During pre-training, POSIT emphasizes two key aspects of substructure identification: firstly, it imposes a soft connectivity constraint to encourage the generation of topologically meaningful substructures; secondly, it aligns resultant substructures with derived prototypes through a prototype-substructure contrastive clustering objective, ensuring attribute-based similarity within clusters. In the fine-tuning stage, a cross-scale attention mechanism is designed to integrate substructure-level information to enhance molecular representations. The effectiveness of the POSIT framework is demonstrated by experimental results from diverse real-world datasets, covering both classification and regression tasks. Moreover, visualization analysis validates the consistency of chemical priors with identified substructures. The source code is publicly available at https://github.com/VRPharmer/POSIT.

The inference of gene regulatory networks (GRNs) is crucial to understanding the regulatory mechanisms that govern biological processes. GRNs may be represented as edges in a graph, and hence, it have been inferred computationally for scRNA-seq data. A wisdom of crowds approach to integrate edges from several GRNs to create one composite GRN has demonstrated improved performance when compared with individual algorithm implementations on bulk RNA-seq and microarray data. In an effort to extend this approach to scRNA-seq data, we present COFFEE (COnsensus single cell-type speciFic inFerence for gEnE regulatory networks), a Borda voting-based consensus algorithm that integrates information from 10 established GRN inference methods. We conclude that COFFEE has improved performance across synthetic, curated, and experimental datasets when compared with baseline methods. Additionally, we show that a modified version of COFFEE can be leveraged to improve performance on newer cell-type specific GRN inference methods. Overall, our results demonstrate that consensus-based methods with pertinent modifications continue to be valuable for GRN inference at the single cell level. While COFFEE is benchmarked on 10 algorithms, it is a flexible strategy that can incorporate any set of GRN inference algorithms according to user preference. A Python implementation of COFFEE may be found on GitHub: https://github.com/lodimk2/coffee.

Plant viral infections cause significant economic losses, totalling $350 billion USD in 2021. With no treatment for virus-infected plants, accurate and efficient diagnosis is crucial to preventing and controlling these diseases. High-throughput sequencing (HTS) enables cost-efficient identification of known and unknown viruses. However, existing diagnostic pipelines face challenges. First, many methods depend on subjectively chosen parameter values, undermining their robustness across various data sources. Second, artifacts (e.g. false peaks) in the mapped sequence data can lead to incorrect diagnostic results. While some methods require manual or subjective verification to address these artifacts, others overlook them entirely, affecting the overall method performance and leading to imprecise or labour-intensive outcomes. To address these challenges, we introduce IIMI, a new automated analysis pipeline using machine learning to diagnose infections from 1583 plant viruses with HTS data. It adopts a data-driven approach for parameter selection, reducing subjectivity, and automatically filters out regions affected by artifacts, thus improving accuracy. Testing with in-house and published data shows IIMI's superiority over existing methods. Besides a prediction model, IIMI also provides resources on plant virus genomes, including annotations of regions prone to artifacts. The method is available as an R package (iimi) on CRAN and will integrate with the web application www.virtool.ca, enhancing accessibility and user convenience.

The functional study of proteins is a critical task in modern biology, playing a pivotal role in understanding the mechanisms of pathogenesis, developing new drugs, and discovering novel drug targets. However, existing computational models for subcellular localization face significant challenges, such as reliance on known Gene Ontology (GO) annotation databases or overlooking the relationship between GO annotations and subcellular localization. To address these issues, we propose DeepMTC, an end-to-end deep learning-based multi-task collaborative training model. DeepMTC integrates the interrelationship between subcellular localization and the functional annotation of proteins, leveraging multi-task collaborative training to eliminate dependence on known GO databases. This strategy gives DeepMTC a distinct advantage in predicting newly discovered proteins without prior functional annotations. First, DeepMTC leverages pre-trained language model with high accuracy to obtain the 3D structure and sequence features of proteins. Additionally, it employs a graph transformer module to encode protein sequence features, addressing the problem of long-range dependencies in graph neural networks. Finally, DeepMTC uses a functional cross-attention mechanism to efficiently combine upstream learned functional features to perform the subcellular localization task. The experimental results demonstrate that DeepMTC outperforms state-of-the-art models in both protein function prediction and subcellular localization. Moreover, interpretability experiments revealed that DeepMTC can accurately identify the key residues and functional domains of proteins, confirming its superior performance. The code and dataset of DeepMTC are freely available at https://github.com/ghli16/DeepMTC.

Stratification of patients diagnosed with cancer has become a major goal in personalized oncology. One important aspect is the accurate prediction of the response to various drugs. It is expected that the molecular characteristics of the cancer cells contain enough information to retrieve specific signatures, allowing for accurate predictions based solely on these multi-omic data. Ideally, these predictions should be explainable to clinicians, in order to be integrated in the patients care. We propose a machine-learning framework based on ensemble learning to integrate multi-omic data and predict sensitivity to an array of commonly used and experimental compounds, including chemotoxic compounds and targeted kinase inhibitors. We trained a set of classifiers on the different parts of our dataset to produce omic-specific signatures, then trained a random forest classifier on these signatures to predict drug responsiveness. We used the Cancer Cell Line Encyclopedia dataset, comprising multi-omic and drug sensitivity measurements for hundreds of cell lines, to build the predictive models, and validated the results using nested cross-validation. Our results show good performance for several compounds (Area under the Receiver-Operating Curve >79%) across the most frequent cancer types. Furthermore, the simplicity of our approach allows to examine which omic layers have a greater importance in the models and identify new putative markers of drug responsiveness. We propose several models based on small subsets of transcriptional markers with the potential to become useful tools in personalized oncology, paving the way for clinicians to use the molecular characteristics of the tumors to predict sensitivity to therapeutic compounds.

The clinical adoption of small interfering RNAs (siRNAs) has prompted the development of various computational strategies for siRNA design, from traditional data analysis to advanced machine learning techniques. However, previous studies have inadequately considered the full complexity of the siRNA silencing mechanism, neglecting critical elements such as siRNA positioning on mRNA, RNA base-pairing probabilities, and RNA-AGO2 interactions, thereby limiting the insight and accuracy of existing models. Here, we introduce siRNADiscovery, a Graph Neural Network (GNN) framework that leverages both non-empirical and empirical rule-based features of siRNA and mRNA to effectively capture the complex dynamics of gene silencing. On multiple internal datasets, siRNADiscovery achieves state-of-the-art performance. Significantly, siRNADiscovery also outperforms existing methodologies in in vitro studies and on an externally validated dataset. Additionally, we develop a new data-splitting methodology that addresses the data leakage issue, a frequently overlooked problem in previous studies, ensuring the robustness and stability of our model under various experimental settings. Through rigorous testing, siRNADiscovery has demonstrated remarkable predictive accuracy and robustness, making significant contributions to the field of gene silencing. Furthermore, our approach to redefining data-splitting standards aims to set new benchmarks for future research in the domain of predictive biological modeling for siRNA.

In the past two decades, genome-wide association studies (GWAS) have pinpointed numerous SNPs linked to human diseases and traits, yet many of these SNPs are in non-coding regions and hard to interpret. Transcriptome-wide association studies (TWAS) integrate GWAS and expression reference panels to identify the associations at gene level with tissue specificity, potentially improving the interpretability. However, the list of individual genes identified from univariate TWAS contains little unifying biological theme, leaving the underlying mechanisms largely elusive. In this paper, we propose a novel multivariate TWAS method that Incorporates Pathway or gene Set information, namely TIPS, to identify genes and pathways most associated with complex polygenic traits. We jointly modeled the imputation and association steps in TWAS, incorporated a sparse group lasso penalty in the model to induce selection at both gene and pathway levels and developed an expectation-maximization algorithm to estimate the parameters for the penalized likelihood. We applied our method to three different complex traits: systolic and diastolic blood pressure, as well as a brain aging biomarker white matter brain age gap in UK Biobank and identified critical biologically relevant pathways and genes associated with these traits. These pathways cannot be detected by traditional univariate TWAS + pathway enrichment analysis approach, showing the power of our model. We also conducted comprehensive simulations with varying heritability levels and genetic architectures and showed our method outperformed other established TWAS methods in feature selection, statistical power, and prediction. The R package that implements TIPS is available at https://github.com/nwang123/TIPS.

Coronary heart disease (CHD) is one of the leading causes of mortality and morbidity in the United States. Accurate time-to-event CHD prediction models with high-dimensional DNA methylation and clinical features may assist with early prediction and intervention strategies. We developed a state-of-the-art deep learning autoencoder survival analysis model (AESurv) to effectively analyze high-dimensional blood DNA methylation features and traditional clinical risk factors by learning low-dimensional representation of participants for time-to-event CHD prediction. We demonstrated the utility of our model in two cohort studies: the Strong Heart Study cohort (SHS), a prospective cohort studying cardiovascular disease and its risk factors among American Indians adults; the Women's Health Initiative (WHI), a prospective cohort study including randomized clinical trials and observational study to improve postmenopausal women's health with one of the main focuses on cardiovascular disease. Our AESurv model effectively learned participant representations in low-dimensional latent space and achieved better model performance (concordance index-C index of 0.864 ± 0.009 and time-to-event mean area under the receiver operating characteristic curve-AUROC of 0.905 ± 0.009) than other survival analysis models (Cox proportional hazard, Cox proportional hazard deep neural network survival analysis, random survival forest, and gradient boosting survival analysis models) in the SHS. We further validated the AESurv model in WHI and also achieved the best model performance. The AESurv model can be used for accurate CHD prediction and assist health care professionals and patients to perform early intervention strategies. We suggest using AESurv model for future time-to-event CHD prediction based on DNA methylation features.