RESUMO
Objective: Information theory has been successfully employed to identify optimal pathway networks, mutual information (MI), and entropy as a dynamic response in statistical methods and estimate input and output information in systems biology. This research aims to investigate potentially integrated gene signatures for bone metastasis using graph-based information theory from the dynamic interaction interphase. Methods: The expression dataset with the series ID GSE26964 for bone metastasis from prostate cancer was retrieved. The dataset was segregated for differentially expressed genes (DEGs) using the Human Cancer Metastasis Database. MI was considered to capture non-linear connections to classify the key DEGs from the collected dataset using gene-gene statistical analysis and then a protein-protein interaction network (PPIN). The PPIN was used to calculate centrality metrics, bottlenecks, and functional annotations. Results: A total of 531 DEGs were identified. Thirteen genes were classified as highly correlated based on their gene expression data matrix. The extended PPIN of the 13 genes comprised 53 nodes and 372 edges. A total of four DEGs were identified as hubs. One novel gene was identified with strong network connectivity. Conclusion: The novel biomarkers for metastasis may provide information on cancer metastasis to the bone by implying MI and information theory.
RESUMO
Objective: The major purpose of the present study was to predict the structure of Radical s-adenosyl-L-methionine Domain 2 (RSAD2), the most targeted protein of the Zika virus using comparative modeling, to validate the models that were generated and molecular dynamics (MD) simulations were performed. Methods: The secondary structure of RSAD2 was estimated using the Garnier-Osguthorpe-Robson, Self-Optimized Prediction method with Alignment, and Position-Specific Iterative-Blast based secondary structure prediction algorithms. The best of them were preferred based on their DOPE score, then three-dimensional structure identification using SWISS-MODEL and the Protein Homology/Analogy Recognition Engine (Phyre2) server. SAVES 6.0 was used to validate the models, and the preferred model was then energetically stabilized. The model with least energy minimization was used for MD simulations using iMODS. Results: The model predicted using SWISS-MODEL was determined as the best among the predicted models. In the Ramachandran plot, there were 238 residues (90.8%) in favored regions, 23 residues (8.8%) in allowed regions, and 1 residue (0.4%) in generously allowed regions. Energy minimization was calculated using Swiss PDB viewer, reporting the SWISS-MODEL with the lowest energy (E = -18439.475 KJ/mol) and it represented a stable structure conformation at three-dimensional level when analyzed by MD simulations. Conclusion: A large amount of sequence and structural data is now available, for tertiary protein structure prediction, hence implying a computational approach in all the aspects becomes an opportunistic strategy. The best three-dimensional structure of RSAD2 was built and was confirmed with energy minimization, secondary structure validation and torsional angles stabilization. This modeled protein is predicted to play a role in the development of drugs against Zika virus infection.
