Structural Comparison of Gene Relevance networks for Breast Cancer Tissues in Different Grades.

Structural Comparison of Gene Relevance networks for Breast Cancer Tissues in Different Grades. Comb Chem High Throughput Screen. 2016 Aug 31; Authors: Zhang Y, Lv K, Dong Y, Su J Abstract In this research, networks of experimental and control groups from gene expression profiles are reversely modeled. By calculating the mutual information between genes, the gene networks are constructed with mutual information values as edge weights under thresholds. Four structural statistics of networks including node degree, clustering coefficient, shortest path length and standard betweenness centrality are used to analyze gene networks for breast cancer tissues in different grades and estrogen receptor (ER) states and comparing them with those for normal breast tissues. Further, the support vector machine (SVM) is used to classify the gene networks into normal and disease states. It is found that the vectors of parameters for normal and disease stages are linearly separable by any combinations of these four structural parameters. From analyzing these network structures, six genes are obtained that may play significant roles in the formation of breast cancer. In our literature review, we find that five out of these six identified genes are closely related to the breast cancer formation as well as its development. This motivates further studies and research on the unidentified gene, which is beneficial to further the diagnosis, treatment and drug...
Source: Combinatorial Chemistry and High Throughput Screening - Category: Chemistry Authors: Tags: Comb Chem High Throughput Screen Source Type: research