Journal of Computational Biology最新文献

Due to the widespread use of antibiotics, many microbes have become drug-resistant. It is urgent to develop new antibiotics that can effectively combat drug-resistant microbes. Exploiting microbe-drug associations can help researchers make progress in drug development. In this paper, we develop for the first time a computational model of Bernoulli random forest (BRF) for microbe-drug association (BRFMDA) prediction. First, we introduced integrated drug similarity and integrated microbe similarity to construct feature of each microbe-drug pair. Second, based on known microbe-drug association, we obtained the features of all positive sample. Then, the same number of negative samples as the number of positive samples were chosen from unknown microbe-drug pairs. Next, we used a filter-based approach to reduce the dimension of features of positive and negative samples. Lastly, BRF was used to train features of positive and negative samples to predict microbe-drug associations. For validating the performance of BRFMDA, we took leave-one-out cross-validation (LOOCV) and fivefold cross-validation, as well as two types of case studies, to validate the prediction performance of BRFMDA. The results of cross-validation and case studies suggested that BRFMDA is a dependable model for predicting potential microbe-drug associations. Specifically, on the Microbe-Drug Association Database (MDAD), BRFMDA obtained an area under the curve (AUC) of 0.9134 in global LOOCV, 0.8958 in local LOOCV, and 0.8657 ± 0.0112 in fivefold cross-validation. On the abiofilm dataset, BRFMDA achieved an AUC of 0.9130 in global LOOCV, 0.8927 in local LOOCV, and 0.8844 ± 0.0137 in fivefold cross-validation.

Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)/Cas9 is a leading genomic editing tool, but its effectiveness is limited by considerable heterogeneity in target efficiency among different single guide RNAs (sgRNA). This study presents RNAS-sgRNA, a hybrid model that integrates neural architecture search (NAS) with recurrent neural networks (RNN) to evaluate the on-target efficacy of CRISPR/Cas9 sgRNA. The RNAS-sgRNA model automates architectural discovery, improving sgRNA sequence categorization without considerable manual adjustment. The NAS component improves the RNN architecture, which analyzes sgRNA sequences represented as binary matrices and produces a classification score. Upon evaluation across several datasets, RNAS-sgRNA exhibits substantial performance enhancements with multiple cell lines, comparing its area under the receiver operating characteristic curve (AUROC) performance to the baseline CRISPRpred(SEQ) and DeepCRISPR models. RNAS-sgRNA demonstrated substantial improvements in AUROC performance in several cell lines compared with existing models. Notable improvements include enhancements of 8.62% for HCT116, 121.57% for HEK293T, 13.40% for HeLa, and 20.78% for HL60 cell lines, resulting in an overall improvement of 13.46%. Compared with DeepCRISPR, the model achieved additional AUROC gains in all cell lines tested, with an average improvement of 14.74%. The study also highlighted the ability of the model to deliver superior performance on smaller datasets through transfer learning, underscoring its potential applications in personalized medicine and genetic research. RNAS-sgRNA introduces a novel integration of NAS with RNN to evaluate the efficacy of CRISPR/Cas9 sgRNA. Unlike traditional methods that require significant manual adjustments, this model automates architectural discovery, optimizing the RNN structure for sgRNA sequence analysis. Furthermore, the application of transfer learning to fine-tune the pretrained model on small cell-line datasets represents a pioneering approach in the domain. The model's demonstrated ability to significantly outperform existing algorithms, including CRISPRpred(SEQ) and DeepCRISPR, across multiple cell lines highlights its innovative contribution to genomic editing research and personalized medicine.

The oral microbiome is a complex environment that consists of diverse microorganisms inhabiting the oral cavity. There are more than 700 different species of bacteria living in the oral cavity which provides nutrition to the microorganisms living in the mouth. As samples tend to be collected with a variation in non-biological factors, batch effects will occur. Batch effects are variations in the same samples, where the variations are affected by the differences in equipment used, the time when the samples were collected, the laboratory conditions, etc. Batch effects can be difficult to address as the variation might not be apparent in individual samples but rather as a whole group between samples. Several research has been proposed to resolve the batch effect, but they tend to require a two-step approach (batch effect removal, and classification), or will suffer from dropout events in gene expressions. In this study, we propose a one-step approach that combines both the batch effect removal and disease classification, eliminating the need for a two-step approach process. LassoNet was used with batch loss to mitigate the effect of batch effect and to classify disease outcome on oral microbiome simultaneously. The model achieved better performance than our baseline models, reaching 0.8 area under the curve on average on the five studies of oral microbiome. In addition, another key aspect of using LassoNet is its ability to carry out feature importance analysis, which is capable to reveal key oral microbiomes associated with disease outcomes.

With the rapid development of single-cell RNA sequencing (scRNA-seq) technology, researchers can now investigate gene expression at the individual cell level. Identifying cell types via unsupervised clustering is a fundamental challenge in analyzing single-cell data. However, due to the high dimensionality of expression profiles, traditional clustering methods often fail to produce satisfactory results. To address this problem, we developed NetworkSSC, a network-guided sparse subspace clustering (SSC) approach. NetworkSSC operates on the same assumption as SSC that cells of the same type have gene expressions lying within the same subspace. In addition, it integrates a regularization term incorporating the gene network's Laplacian matrix, which captures functional associations between genes. Comparative analysis on nine scRNA-seq datasets shows that NetworkSSC outperforms traditional SSC and other unsupervised methods in most cases.

Elucidating the functional mechanisms underlying most associations between phenomes and genomes uncovered by genome-wide association studies remains a challenging problem. Deep neural networks that excel in feature learning from sequential data have recently emerged as promising approaches to addressing this challenge by mapping sequence patterns in DNA to functional genomic events. Despite the impressive progress made in this regard, the existing studies are largely limited to examining a type of network architecture that primarily consists of simple stacked convolutional layers of filters of a uniform size. These networks lack the consideration of specifics in the mapping of DNA sequences to functional genomic events, thereby impairing the learning efficiency of these networks. To address this problem, in this article, we propose an efficient DNA sequence learner (EDSL), a novel biologically informed architecture that (1) introduces filters of varying sizes in the first convolutional layer to enhance the learning of sequence patterns of diverse sizes and (2) utilizes dense connections to facilitate the participation of sequence patterns at varying levels in prediction. Our results regarding both synthetic data and a dataset consisting of 367 experimentally derived functional genomic profiles demonstrate the effectiveness of the proposed design choices and the superiority of the EDSL over existing networks in terms of both prediction performance and sequence pattern learning. Moreover, our ablation study indicates that both the proposed design choices enhance learning-importantly, in a differential and complementary manner.

Protein abnormalities disrupt various cellular and contribute to disease development. Identifying disease-associated proteins is crucial for precision medicine, but traditional methods are time-consuming and costly, necessitating computational approaches. Existing computational methods rely on manual feature engineering and fail to leverage deep features from amino acid sequences and protein structures. In this article, we propose Model for predicting protein-phenotype associations by Fusing multi-view Features (MFF-HPO), a model for predicting protein-phenotype associations by fusing multi-view features from amino acid sequences. First, we generate three-dimensional protein structure from amino acid sequence to derive contact graphs and secondary structures then integrate these with direct sequence encoding and physicochemical properties. Using a Graph Attention Network, we extract structural features from contact graphs, while deep neural networks capture global and local features from secondary structures, physicochemical properties, and sequence encoding. Finally, concatenated features are used to predict phenotype annotations. MFF-HPO outperforms state-of-the-art methods with a mean area under the precision-recall curve of 0.314 and a mean F_max of 0.371. Ablation studies confirm that multi-view feature fusion enhances predictions, and case studies validate its practicality.

Radiology report generation (RRG) tasks leverage computer-aided technology to automatically produce descriptive text reports for medical images, aiming to ease radiologists' workload, reduce misdiagnosis rates, and lessen the pressure on medical resources. However, previous works have yet to focus on enhancing feature extraction of low-quality images, incorporating cross-modal interaction information, and mitigating latency in report generation. We propose an Image-Enhanced Cross-Modal Fusion Network (IFNet) for automatic RRG to tackle these challenges. IFNet includes three key components. First, the image enhancement module enhances the detailed representation of typical and atypical structures in X-ray images, thereby boosting detection success rates. Second, the cross-modal fusion networks efficiently and comprehensively capture the interactions of cross-modal features. Finally, a more efficient transformer report generation module is designed to optimize report generation efficiency while being suitable for low-resource devices. Experimental results on public datasets IU X-ray and MIMIC-CXR demonstrate that IFNet significantly outperforms the current state-of-the-art methods.

Cervical cancer is the fourth most common disease among women worldwide, and pap smear images are used as a primary diagnostic technique to detect precancerous and cancerous abnormalities in the cervix, vagina, and vulva. Deep learning algorithms have gained popularity in developing automated computer-aided diagnostic systems to solve the difficulties associated with manual assessment. This article introduces an innovative hybrid approach to effectively and accurately categorizing cervical cells. The proposed model employs advanced data enhancement techniques, including resampling to address class imbalance and augmentation (e.g., random horizontal flips and rotations) to increase dataset diversity and improve generalization. These strategies help the model handle different types of data more effectively, making it more adaptable and reliable in real-world scenarios. We use Vision Transformer's (ViT) linear projection and position embedding to change the input images into patches that can be sent to a transformer encoder. A fusion architecture is established by incorporating supplementary convolutional layers, followed by a fully connected layer, to improve the features extracted by the model. The ViT-based model is developed using pretrained weights and allows fine-tuning to address problems with cervical cancer classification efficiently. To enhance the quality of these cell images, we employ median smoothing and Gaussian filtering as preprocessing techniques. The experiment results demonstrate the proposed methodology's potential for improving the precision of cervical cancer classification. Notably, our model exhibited outstanding accuracy on the 2-state classification on the Herlev dataset and the 3-state classification on the SIPaKMeD dataset, at 98.07% and 98.08%, respectively. The model's ability to effectively categorize cervical cancer images across various datasets is evidenced by the accuracy rates specific to each dataset. This indicates the model's robustness and promise for practical clinical use.