Prediction of Escherichia Coli K-12 Promoters Using Convolutional Neural Network
Issue:
Volume 6, Issue 2, December 2018
Pages:
31-35
Received:
11 October 2018
Accepted:
31 October 2018
Published:
30 November 2018
DOI:
10.11648/j.cbb.20180602.11
Downloads:
Views:
Abstract: Promoters are significant cis-acting elements in genomes and play important roles in gene regulation. Each gene is regulated by a specific type of promoter, so determining the type of promoter for regulation of a gene is crucial to explore the gene function. Although some computational methods to predict promoters have been proposed, their performances are not satisfying. Convolutional neural network (CNN) is a powerful model in deep learning, it has been applied in bioinformatics in recent years. To improve the performance of promoter prediction, in this study, six types of Escherichia coli K-12 promoter DNA sequences were collected from the RegulonDB database, and constructed a CNN model to predict promoters using the Keras platform. The CNN model is composed of two convolutional layers, three dropout layers, four batch normalization layers and one hidden layer. To evaluate the performances of the CNN model, the 10-fold cross-validation and the receiver operating characteristic (ROC) curve plotting were performed. The results show, the accuracies of predictions for promoters sigma 24, sigma 28, sigma 32, sigma 38, sigma 54 and sigma 70 are 94%, 97%, 95%, 95%, 97% and 83%, respectively. The convolutional neural network model achieves the highest accuracy in promoter prediction up to now. In conclusion, CNN is the best model in promoter prediction, and it will be a promising model both in DNA and protein sequence analysis.
Abstract: Promoters are significant cis-acting elements in genomes and play important roles in gene regulation. Each gene is regulated by a specific type of promoter, so determining the type of promoter for regulation of a gene is crucial to explore the gene function. Although some computational methods to predict promoters have been proposed, their performa...
Show More
Docking Studies for Assessment of Wound Healing Potential of Dalethyne Derivatives: An in Silico Approach
Issue:
Volume 6, Issue 2, December 2018
Pages:
36-51
Received:
1 October 2018
Accepted:
19 November 2018
Published:
18 December 2018
DOI:
10.11648/j.cbb.20180602.12
Downloads:
Views:
Abstract: Background: A cascade of enzymes acting in union is involved in the natural wound healing pharmacology of humans making the process a lengthy one. This in turns necessitates new synthetic molecules effective in accelerating the wound healing process. Objective: The present work deals with synthetic molecules aimed at healing wounds targeting the essential enzymes involved in the wound healing process. Method: A series of in house synthesized dalethyne derivatives have been studied in the present work based on their ability to interact with the requisite proteins using docking methodology and degree of interactions of the molecules with each of the proteins have been determined based on their binding energy values. Subsequently, the inhibitory concentrations of the molecules were also predicted based of docking statistics. The validation of the procedure was performed based on the docking interactions of the native ligand. Results: The dalethyne derivatives showed effective interactions with the amino acid residues present in the active site of some of the essential proteins involved in the wound healing process accounting for the conducive effects of these molecules in the wound healing process. Conclusion: The present work thus provides a meaningful insight as to the structural requirements of the dalethyne derivatives that would facilitate their interaction with the receptors involved in the wound healing process such that the molecules can be efficiently formulated into a pharmaceutical dosage form.
Abstract: Background: A cascade of enzymes acting in union is involved in the natural wound healing pharmacology of humans making the process a lengthy one. This in turns necessitates new synthetic molecules effective in accelerating the wound healing process. Objective: The present work deals with synthetic molecules aimed at healing wounds targeting the es...
Show More