China
Africa
Explore chapters and articles related to this topic
Insight into Knapsack Metabolite Ecology Database: A Comprehensive Source of Species: Voc-Biological Activity Relationships
Published in Raquel Cumeras, Xavier Correig, Volatile organic compound analysis in biomedical diagnosis applications, 2018
Azian Azamimi Abdullah, M.D. Altaf-Ul-Amin, Shigehiko Kanaya
There are many VOCs, which have several biological activities. Thus, it is important to investigate the relationships between VOCs and their biological activities statistically. Initially, we determined the pairwise chemical structural similarity between VOCs based on Tanimoto coefficient. 2-D compound structures in the generic structure definition file (SDF) format of all 341 VOCs were obtained from PubChem database and then, were imported into ChemmineR package in one batch file. We calculated the chemical structure similarity using Tanimoto coefficient. Then, we converted the Tanimoto similarity matrix into distance matrix by subtracting each of the similarity values from 1. Based on distance matrix, we performed heatmap clustering, and the result is shown in Figure 9.9. White and red colors indicate the extreme distance values of 0 and 1 respectively and the intermediate distance values are indicated by the intensity of the red color. From the heatmap plot, we tentatively outlined 11 clusters of VOCs. The count of VOCs belonging to each activity group in each cluster is shown in Table 9.1. To assess the richness of VOCs of similar activity in individual clusters, we determined their p-values based on hypergeometric distribution, which are also shown in Table 9.1. The major types of chemical compounds belonging to each cluster and their corresponding biological activities are mentioned in Table 9.2. The detail description of chemical compounds in each cluster and their related biological activities can be found in (Abdullah et al., 2015). Additionally, we also compared several types of hierarchical clustering methods (single, complete, average, centroid, median linkage and Ward’s method) with DPClus algorithm to cluster the chemical structures of VOCs using Tanimoto coefficient as a similarity measure. We found that Ward’s method has the most matching clusters with DPClus while median has the least matching clusters (Abdullah et al., 2016). Compared to hierarchical clustering, DPClus can give a better visualization of how generated clusters interact with each other, and we found that VOCs belonging to the interacted clusters have similar chemical structure, which indicates possibilities of exhibiting similar biological activities. Both clustering methods show that there are strong links between the chemical structure of VOCs and their biological activities. Comparative activity relationships between chemical ecology and human health care activity will lead to the systematization of metabolomics combined with human and ecological metabolic pathways.
Experimental data-based reduced-order model for analysis and prediction of flame transition in gas turbine combustors
Published in Combustion Theory and Modelling, 2019
Shivam Barwey, Malik Hassanaly, Qiang An, Venkat Raman, Adam Steinberg
Centroid groups are further identified via the cluster distance matrix shown in Figure 3(b), which displays the L2 distance between each centroid [19]. The distance matrix is symmetric with diagonal equal to zero.