Journal of Bioinformatics and Computational Biology最新文献

Lung adenocarcinoma (LUAD) seriously threatens human health and generally results from dysfunction of relevant module molecules, which dynamically change with time and conditions, rather than that of an individual molecule. In this study, a novel network construction algorithm for identifying early warning network signals (IEWNS) is proposed for improving the performance of LUAD early diagnosis. To this end, we theoretically derived a dynamic criterion, namely, the relationship of variation (RV), to construct dynamic networks. RV infers correlation [Formula: see text] statistics to measure dynamic changes in molecular relationships during the process of disease development. Based on the dynamic networks constructed by IEWNS, network warning signals used to represent the occurrence of LUAD deterioration can be defined without human intervention. IEWNS was employed to perform a comprehensive analysis of gene expression profiles of LUAD from The Cancer Genome Atlas (TCGA) database and the Gene Expression Omnibus (GEO) database. The experimental results suggest that the potential biomarkers selected by IEWNS can facilitate a better understanding of pathogenetic mechanisms and help to achieve effective early diagnosis of LUAD. In conclusion, IEWNS provides novel insight into the initiation and progression of LUAD and helps to define prospective biomarkers for assessing disease deterioration.

Gel zymography quantifies the activity of certain enzymes in tumor processes. These enzymes are widely used in medical diagnosis. In order to analyze them, experts classify the zymography spots into various classes according to their tonalities. This classification is done by visual analysis, which is what makes it a subjective process. This work proposes a methodology to carry out this classifications with a process that involves an unsupervised learning algorithm in the images, denoted as the GI algorithm. With the experiments shown in this paper, this methodology could constitute a tool that bioinformatics scientists can trust to perform the desired classification since it is a quantitative indicator to order the enzymatic activity of the spots in a zymography.

In this study, we propose Feedback-AVPGAN, a system that aims to computationally generate novel antiviral peptides (AVPs). This system relies on the key premise of the Generative Adversarial Network (GAN) model and the Feedback method. GAN, a generative modeling approach that uses deep learning methods, comprises a generator and a discriminator. The generator is used to generate peptides; the generated proteins are fed to the discriminator to distinguish between the AVPs and non-AVPs. The original GAN design uses actual data to train the discriminator. However, not many AVPs have been experimentally obtained. To solve this problem, we used the Feedback method to allow the discriminator to learn from the existing as well as generated synthetic data. We implemented this method using a classifier module that classifies each peptide sequence generated by the GAN generator as AVP or non-AVP. The classifier uses the transformer network and achieves high classification accuracy. This mechanism enables the efficient generation of peptides with a high probability of exhibiting antiviral activity. Using the Feedback method, we evaluated various algorithms and their performance. Moreover, we modeled the structure of the generated peptides using AlphaFold2 and determined the peptides having similar physicochemical properties and structures to those of known AVPs, although with different sequences.

Glycoproteins play an important and ubiquitous role in many biological processes such as protein folding, cell-to-cell signaling, invading microorganism infection, tumor metastasis, and leukocyte trafficking. The key mechanism of glycoproteins must be revealed to model and refine glycosylated protein recognition, which will eventually assist in the design and discovery of carbohydrate-derived therapeutics. Experimental procedures involving wet-lab experiments to reveal glycoproteins are very time-consuming, laborious, and highly costly. However, costly and tedious experimental procedures can be assisted by ranking the most probable glycoproteins through computational methods with improved accuracy. In this study, we have proposed a novel machine learning-based predictive model for glycoproteins identification. Our proposed model is based on sequence-derived structural descriptors (SDSD) that fill the gap of unavailability of protein 3D structures and lack of accuracy in sequence information alone. Through a series of simulation studies, we have shown that our proposed model gives state-of-the-art generalization performance verified through various machine learning-centric and biologically relevant techniques and metrics. Through data mining in this study, we have also identified the role of descriptors in determining glycoproteins. Python-based standalone code together with a webserver implementation of our proposed model (COYOTE: identifiCation Of glYcoprOteins Through sEquences) is available at the URL: https://sites.google.com/view/wajidarshad/software.

Identification of potential drug-target interactions (DTIs) plays a pivotal role in the development of drug and target discovery in the public healthcare sector. However, biological experiments for predicting interactions between drugs and targets are still expensive, complicated, and time-consuming. Thus, computational methods are widely applied for aiding drug-target interaction prediction. In this paper, we propose a novel model, named GCMCDTI, for DTIs prediction which adopts a graph convolutional network based on matrix completion. We regard the association prediction between drugs and targets as link prediction and treat the process as matrix completion, and then a graph convolutional auto-encoder framework is employed to construct the drug and target embeddings. Then, a bilinear decoder is applied to reconstruct the DTI matrix. We conduct our experiments on four benchmark datasets consisting of enzymes, G protein-coupled receptors (GPCRs), ion channels, and nuclear receptors. The five-fold cross-validation results achieve the high average AUC values of 95.78%, 95.31%, 93.90%, and 91.77%, respectively. To further evaluate our method, we compare our proposed method with other state-of-the-art approaches. The comparison results illustrate that our proposed method obtains improvement in performance on DTI prediction. The proposed method will be a good choice in the field of DTI prediction.

We present hybrid system-based gene regulatory network models for lambda, HK022, and Mu bacteriophages together with dynamics analysis of the modeled networks. The proposed lambda phage model LPH2 is based on an earlier work and incorporates more recent biological assumptions about the underlying gene regulatory mechanism, HK022, and Mu phage models are new. All three models provide accurate representations of experimentally observed lytic and lysogenic behavioral cycles. Importantly, the models also imply that lysis and lysogeny are the only stable behaviors that can occur in the modeled networks. In addition, the models allow to derive switching conditions that irrevocably lead to either lytic or lysogenic behavioral cycle as well as constraints that are required for their biological feasibility. For LPH2 model the feasibility constraints place two mutually independent requirements on comparative order of cro and cI protein binding site affinities. However, HK022 model, while broadly similar, does not require any of these constraints. Biologically very different lysis-lysogeny switching mechanism of Mu phage is also accurately reproduced by its model. In general the results show that hybrid system model (HSM) hybrid system framework can be successfully applied to modeling small ([Formula: see text] gene) regulatory networks and used for comprehensive analysis of model dynamics and stable behavior regions.

The profound impact of in silico studies for a fast-paced drug discovery pipeline is undeniable for pharmaceutical community. The rational design of novel drug candidates necessitates considering optimization of their different aspects prior to synthesis and biological evaluations. The affinity prediction of small ligands to target of interest for rank-ordering the potential ligands is one of the most routinely used steps in the context of virtual screening. So, the end-point methods were employed for binding free energy estimation focusing on evaluating simulation time effect. Then, a set of human aldose reductase inhibitors were selected for molecular dynamics (MD)-based binding free energy calculations. A total of 100[Formula: see text]ns MD simulation time was conducted for the ligand-receptor complexes followed by prediction of binding free energies using MM/PB(GB)SA and LIE approaches under different simulation time. The results revealed that a maximum of 30[Formula: see text]ns simulation time is sufficient for determination of binding affinities inferred from steady trend of squared correlation values (R²) between experimental and predicted [Formula: see text]G as a function of MD simulation time. In conclusion, the MM/PB(GB)SA algorithms performed well in terms of binding affinity prediction compared to LIE approach. The results provide new insights for large-scale applications of such predictions in an affordable computational cost.