Computational Biology and Chemistry最新文献

Protein lysine crotonylation is an important post-translational modification that regulates various cellular activities. For example, histone crotonylation affects chromatin structure and promotes histone replacement. Identification and understanding of lysine crotonylation sites is crucial in the field of protein research. However, due to the increasing amount of non-histone crotonylation sites, existing classifiers based on traditional machine learning may encounter performance limitations. In order to address this problem, a novel deep learning-based model for identifying crotonylation sites is presented in this study, given the unique advantages of deep learning techniques for sequence data analysis. In this study, an MLP-Attention-based model was developed for the identification of crotonylation sites. Firstly, three feature extraction strategies, namely Amino Acid Composition, K-mer, and Distance-based residue features extraction strategy, were used to encode crotonylated and non-crotonylated sequences. Then, in order to balance the training dataset, the FCM-GRNN undersampling algorithm combining fuzzy clustering and generalized neural network approaches was introduced. Finally, to improve the effectiveness of crotonylation site identification, we explored various classification algorithms, and based on the relevant experimental performance comparisons, the multilayer perceptron (MLP) combined with the superimposed self-attention mechanism was finally selected to construct the prediction model ILYCROsite. The results obtained from independent testing and five-fold cross-validation demonstrated that the model proposed in this study, ILYCROsite, had excellent performance. Notably, on the independent test set, ILYCROsite achieves an AUC value of 87.93 %, which is significantly better than the existing state-of-the-art models. In addition, SHAP (Shapley Additive exPlanations) values were used to analyze the importance of features and their impact on model predictions. Meanwhile, in order to facilitate researchers to use the prediction model constructed in this study, we developed a prediction program to identify the crotonylation sites in a given protein sequence. The data and code for this program are available at: https://github.com/wmqskr/ILYCROsite.

A potent growth inhibitor for normal mammary epithelial cells is transforming growth factor beta 1 (TGF-β1). When breast tissues lose the anti-proliferative activity of this factor, invasion and bone metastases increase. Human breast cancer (hBC) cells express more activating transcription factor 3 (ATF3) when exposed to TGF-β1, and this transcription factor is essential for BC development and bone metastases. Non-coding RNAs (ncRNAs), including circular RNAs (circRNAs) and microRNAs (miRNAs), have emerged as key regulators controlling several cellular processes. In hBC cells, TGF-β1 stimulated the expression of hsa-miR-4653–5p that putatively targets ATF3. Bioinformatics analysis predicted that hsa-miR-4653–5p targets several key signaling components and transcription factors, including NFKB1, STAT1, STAT3, NOTCH1, JUN, TCF3, p300, NRF2, SUMO2, and NANOG, suggesting the diversified role of hsa-miR-4653–5p under physiological and pathological conditions. Despite the high abundance of hsa-miR-4653–5p in hBC cells, the ATF3 level remained elevated, indicating other ncRNAs could inhibit hsa-miR-4653–5p’s activity. In silico analysis identified several circRNAs having the binding sites for hsa-miR-4653–5p, indicating the sponging activity of circRNAs towards hsa-miR-4653–5p. The study's findings suggest that TGF-β1 regulates circRNAs and hsa-miR-4653–5p, which in turn affects ATF3 expression, thus influencing BC progression and bone metastasis. Therefore, focusing on the TGF-β1/circRNAs/hsa-miR-4653–5p/ATF3 network could lead to new ways of diagnosing and treating BC.

Amyotrophic lateral sclerosis (ALS), commonly known as Lou Gehrig's disease, is a debilitating neurodegenerative disorder characterized by the progressive degeneration of nerve cells in the brain and spinal cord. Despite extensive research, its precise etiology remains elusive, and early diagnosis is challenging due to the absence of specific tests. This study aimed to identify potential blood-based biomarkers for early ALS detection and monitoring using datasets from whole blood samples (GSE112680) and oligodendrocytes, astrocytes, and fibroblasts (GSE87385) obtained from the NCBI-GEO repository. Through bioinformatics analysis, including protein-protein interactions and molecular pathway analyses, we identified differentially expressed genes (DEGs) associated with ALS. Notably, ALS2, ADH7, ALDH8A1, ALDH3B1, ABHD2, ABHD17B, ABHD12, ABHD13, PGAM2, AURKB, ANAPC11, VAPA, UNC45B, and TNNT2 emerged as top-ranked DEGs, implicated in drug metabolism, protein depalmytilation, and the AKT/mTOR signaling pathways. Among these, AurKB established as a potential therapeutic biomarker with relevance to various neurological conditions. Consequently, AurKB was selected for identifying potential therapeutic molecules and utilized for in silico structural characterization studies. Exploration of the IMPATT database led to the discovery of a lead compound similar to Fostamatinib, currently used for AurKB. Initial molecular docking and MMGBSA-based binding energy analysis were followed by molecular dynamics simulation (MDS) and free energy landscape (FEL) analysis to validate the ligand's binding efficacy and understand dynamic processes within the biological system. The identified potential biomarkers and lead molecule provide novel insights into the correlation between blood cell transcripts and ALS pathology, paving the way for blood-based diagnostic tools for early ALS detection and ongoing disease monitoring.

Apoptotic proteins play a crucial role in the apoptosis process, ensuring a balance between cell proliferation and death. Thus, further elucidating the regulatory mechanisms of apoptosis will enhance our understanding of their functions. However, the development of computational methods to accurately identify positive and negative regulation of apoptosis remains a significant challenge. This work proposes a machine learning model based on multi-feature fusion to effectively identify the roles of positive and negative regulation of apoptosis. Initially, we constructed a reliable benchmark dataset containing 200 positive regulation of apoptosis and 241 negative regulation of apoptosis proteins. Subsequently, we developed a classifier that combines the support vector machine (SVM) with pseudo composition of k-spaced amino acid pairs (PseCKSAAP), composition transition distribution (CTD), dipeptide deviation from expected mean (DDE), and PSSM-composition to identify these proteins. Analysis of variance (ANOVA) was employed to select optimized features that could yield the maximum prediction performance. Evaluating the proposed model on independent data revealed and achieved an accuracy of 0.781 with an AUROC of 0.837, demonstrating our model's potent capabilities.

The article explores the polypharmacological profiling of 4-((5-(decylthio)-4-methyl-4H-1,2,4-triazole-3-yl)methyl)morpholine as a potential antimicrobial agent. The study utilized 15148 electronic pharmacophore models of organisms, ranked by the Tversky index. Detailed analysis revealed classical bonding patterns with selected enzymes, identifying key amino acid residues involved in complex formation. Protein target prediction was conducted through various stages using the Galaxy web service, including ligand structure creation, pharmacophore alignment, and target ranking. The activities of the molecules against 1G6C, 2W6O, 3G7F, 3OWU, 4IVR, and 4TZT proteins were compared. Docking studies with PyMOL and Discovery Studio Visualizer revealed binding to thymidine kinase, thiamine phosphate synthase, and biotin carboxylase with promising binding affinities. These interactions suggest potential antibacterial and antiviral effects, warranting further virtual screening and in-depth studies for the development of effective antimicrobial drugs. Calculations of the molecules were made with the gaussian package program. Calculations were made on the 6-31++g** basis set at B3LYP, HF, and M062X levels with Gaussian software. Afterwards, the 0–100 ns interaction of the molecule with the highest activity was examined.

Fusion proteins have the potential to become the new norm for targeted therapeutic treatments. Highly specific payload delivery can be achieved by combining custom targeting moieties, such as V_HH domains, with active parts of proteins that have a particular activity not naturally targeted to the intended cells. Conversely, novel drug products may make use of the highly specific targeting properties of naturally occurring proteins and combine them with custom payloads. When designing such a product, there is rarely a known structure for the final construct which makes it difficult to assess molecular behaviour that may ultimately impact therapeutic outcome. Considering the time and cost of expressing a construct, optimising the purification procedure, obtaining sufficient quantities for biophysical characterisation, and performing structural studies in vitro, there is an enormous benefit to conduct in silico studies ahead of wet lab work.

By following a repeatable, streamlined, and fast workflow of molecular dynamics assessment, it is possible to eliminate low-performing candidates from costly experimental work. There are, however, many aspects to consider when designing a novel fusion protein and it is crucial not to overlook some elements. In this work, we suggest a set of user-friendly, open-source methods which can be used to screen fusion protein candidates from the sequence alone. We used the light chain and translocation domain of botulinum toxin A (BoNT/A) fused with a selected V_HH domain, termed here LC-H_N-V_HH, as a case study for a general approach to designing, modelling, and simulating fusion proteins. Its behaviour in silico correlated well with initial in vitro work, with SEC HPLC showing multiple protein states in solution and a dynamic protein shifting between these states over time without loss of material.

A medicinal pteridophyte known as Attukal Kizhangu L. has been used to cure patients for centuries by administering plant parts based on conventional and common practices. Regarding its biological functions, significant use and advancement have been made. Extract of Attukal Kizhangu L. is the subject of the current study, which uses network pharmacology as its foundation. Three targeted compounds such as α-Lapachone, Dihydrochalcone, and Piperine were chosen for additional research from the 17 Phytoconstituents that were filtered out by the Coupled UPLC-HRMS study since they followed to Lipinski rule and showed no toxicity. The pharmacokinetics and physicochemical properties of these targeted compounds were analyzed by using three online web servers pkCSM, Swiss ADME, and Protox-II. This is the first in silico study to document these compound's effectiveness against the standard drug DOX in treating Periodontitis. The Swiss target prediction database was used to retrieve the targets of these compounds. DisGeNET and GeneCards were used to extract the targets of periodontitis. The top five hub genes were identified by Cytoscape utilizing the protein-protein interaction of common genes, from which two hub genes and three binding proteins of collagenase enzymes were used for further studies AA2, PGE2, PI2, TNFA, and PGP. The minimal binding energy observed in molecular docking, indicative of the optimal docking score, corresponds to the highest affinity between the protein and ligand. To corroborate the findings of the docking study, molecular dynamics (MD) simulations, and MMPBSA calculations were conducted for the complexes involving AA2-α-LPHE, AA2-DHC, and AA2-PPR. This research concluded that AA2-DHC was the most stable complex among the investigated interactions, surpassing the stability of the other complexes examined in comparison with the standard drug DOX. Overall, the findings supported the promotion of widespread use of Attukal Kizhangu L. in clinics as a potential therapeutic agent or may be employed for the treatment of acute and chronic Periodontitis.

The tertiary lymphoid structure (TLS) plays a central role in cancer immune response, and its gene expression pattern, called the TLS signature, has shown prognostic value in breast cancer. The formation of TLS and tumor-associated high endothelial venules (TA-HEVs), responsible for lymphocytic infiltration within the TLS, is associated with the expression of cancer hallmark genes (CHGs) related to immunity and inflammation. In this study, we performed co-expression network analysis of immune- and inflammation-related CHGs to identify predictive genes for breast cancer. In total, 382 immune- and inflammation-related CHGs with high expression variance were extracted from the GSE86166 microarray dataset of patients with breast cancer. CHGs were classified into five modules by applying weighted gene co-expression network analysis. The survival analysis results for each module showed that one module comprising 45 genes was statistically significant for relapse-free and overall survival. Four network properties identified key genes in this module with high prognostic prediction abilities: CD34, CXCL12, F2RL2, JAM2, PROS1, RAPGEF3, and SELP. The prognostic accuracy of the seven genes in breast cancer was synergistic and exceeded that of other predictors in both small and large public datasets. Enrichment analysis predicted that these genes had functions related to leukocyte infiltration of TA-HEVs. There was a positive correlation between key gene expression and the TLS signature, suggesting that gene expression levels are associated with TLS density. Co-expression network analysis of inflammation- and immune-related CHGs allowed us to identify genes that share a standard function in cancer immunity and have a high prognostic predictive value. This analytical approach may contribute to the identification of prognostic genes in TLS.

In the realm of plant biology, understanding the intricate regulatory mechanisms governing stress responses stands as a pivotal pursuit. Circular RNAs (circRNAs), emerging as critical players in gene regulation, have garnered attention in recent days for their potential roles in abiotic stress adaptation. A comprehensive grasp of circRNAs' functions in stress response offers avenues for breeders to manipulating plants to develop abiotic stress resistant crop cultivars to thrive in challenging climates. This study pioneers a machine learning-based model for predicting abiotic stress-responsive circRNAs. The K-tuple nucleotide composition (KNC) and Pseudo KNC (PKNC) features were utilized to numerically represent circRNAs. Three different feature selection strategies were employed to select relevant and non-redundant features. Eight shallow and four deep learning algorithms were evaluated to build the final predictive model. Following five-fold cross-validation process, XGBoost learning algorithm demonstrated superior performance with LightGBM-chosen 260 KNC features (Accuracy: 74.55 %, auROC: 81.23 %, auPRC: 76.52 %) and 160 PKNC features (Accuracy: 74.32 %, auROC: 81.04 %, auPRC: 76.43 %), over other combinations of learning algorithms and feature selection techniques. Further, the robustness of the developed models were evaluated using an independent test dataset, where the overall accuracy, auROC and auPRC were found to be 73.13 %, 72.34 % and 72.68 % for KNC feature set and 73.52 %, 79.53 % and 73.09 % for PKNC feature set, respectively. This computational approach was also integrated into an online prediction tool, AScirRNA (https://iasri-sg.icar.gov.in/ascirna/) for easy prediction by the users. Both the proposed model and the developed tool are poised to augment ongoing efforts in identifying stress-responsive circRNAs in plants.

Numerous studies have shown that microRNAs (miRNAs) play a key role in human diseases as critical biomarkers. Its abnormal expression is often accompanied by the emergence of specific diseases. Therefore, studying the relationship between miRNAs and diseases can deepen the insights of their pathogenesis, grasp the process of disease onset and development, and promote drug research of specific diseases. However, many undiscovered relationships between miRNAs and diseases remain, significantly limiting research on miRNA-disease correlations. To explore more potential correlations, we propose a dual-channel convolutional model based on stacked deep learning collaborative gradient decomposition for predicting miRNA-disease associations (DCSGMDA). Firstly, we constructed similarity networks for miRNAs and diseases, as well as an association relationship network. Secondly, potential features were fully mined using stacked deep learning and gradient decomposition networks, along with dual-channel convolutional neural networks. Finally, correlations were scored by a multilayer perceptron. We performed 5-fold and 10-fold cross-validation experiments on DCSGMDA using two datasets based on the Human MicroRNA Disease Database (HMDD). Additionally, parametric, ablation, and comparative experiments, along with case studies, were conducted. The experimental results demonstrate that DCSGMDA performs well in predicting miRNA-disease associations.