Using Different Approaches For Identification And Investigation Diseases Based On Genetic Data
Professor: Luo Jiawei
Computer Science & Technology
College of computer science and Electronic Engineering
The development of new approaches and algorithms is substantial to analyze the behavior of genes and of interactions between them. Study of gene expression and the associated regulation mechanisms, is the challenge to obtaining an integrated analysis that reflects genetic activity in the organism. Describe the complex interactions among genes with nonlinear property and investigate genes, leads to important results contribute to know how the diseases formation. Transcription factors (TFs) and microRNAs (miRNAs) often are found together in a networks at the transcriptional and post-transcriptional level that is lead to a combinatorial and complex regulation of biological processes. We have applied three datasets of cancer are obtained from the TCGA project webpage, gastric cancer, breast cancer and thyroid. In this research, we proposed three methods to answer three main questions: - How to identification gastric cancer related genes from Deep Sequencing Data? - How to composition combinatorial regulatory network from TFs and microRNA factors? And - how to analyzing the complex interactions in biological networks based on based on the available biological tools?
For the first question, we proposed a method to measurement the nonlinear association distance between two random variables based on se the distance correlation. In addition to utilizing weighted gene Co-expression network analysis (WGCNA) of gene expression data. The significantly differentially expressed genes detected with the DESeq method are used to construct the co-expression network. For the second question, initially, we used the k-means cluster and ANOVA test to filtering and selection the significant genes. Then, the combinatorial regulatory network composed of TFs and miRNAs regulations is constructed through the forward engineering strategy to searching of TFs binding sites or miRNA seed regions in the putative target sequences. For the third question, we discussed in depth how create and visualize the biological network in Cytoscape based on Nested network format (NNF) and the Boolean Meta-Filter to search of subsets nodes or edges based on specific conditions. Then, we used the KeyPathwayMiner as heuristics method, to find pathways related to cancer disease and extract the sub-networks.
In the first approach with gastric genes, the results show some genes are enriched in some biological processes and pathways, including cell cycle, mitosis and chromosome segregation, DNA replication and p53 pathway. In addition to, the connectivity of all the genes appear the circumstance of loss in the gastric cancer network, especially to the absent genes. These results indicate that the activation of these genes may play important roles in the gastric cancer. In the second approach, our gene regulatory network (GRN), uncovered some of the regulators like CREB1, E2F1, hsa-miR-106b, hsa-mir-200c AHR and ARNT that play significant roles in the regulation of cell proliferation, such as A549 cells. The combinatorial gene regulatory network included several of TFs and microRNA regulations, permitting the investigation and study in further from the genes that are related to cancer. We evaluated the gene networks for breast cancer and thyroid cancer, by computed the area under the receiver operating characteristic curve (AUC) and draw the ROC curve. We generated a ranking of regulatory relationship based on importance score of relationship. The AUC values are 0.95 and 0.91 for breast cancer and thyroid cancer networks respectively. In the third approach, our results found CCNB1, SP1 AHR and MIA3 genes participated in some biological processes related to cancerous diseases, such as DNA Replication, the secular response to stress, Postsynaptic density and Asymmetric synapse. Which indicated that these genes also had a role in the formation of cancerous diseases.
The main goal of this work aims to build a mathematical model that can contribute effectively to the detection of the genes involved in the formation of cancerous diseases. Also, this work contributes to enriching the subject of personalized medicine that is aiming for a therapeutic approach involving the use of an individual's genetic and epigenetic information to tailor drug therapy or preventive care.
Hesham Abdulatef Mohammed Al-bukhaiti, Jiawei Luo, Using differential nonlinear gene co-expression network for identification gastric cancer-related genes. Biomedical Research, 2017, 28(16), 1-4, (SCI).
Hesham Abdulatef Mohammed Al-bukhaiti, Jiawei Luo, A hybrid approach to select and Combine Gene Expression Regulation based on cancer datasets, IOP Conference. Series: Materials Science and Engineering 2020, (EI Compendex)
版权所有?湖南大学2017 湖南大学党委宣传部 地址：湖南省长沙市岳麓区麓山南路麓山门 邮编：410082 Email：firstname.lastname@example.org 域名备案信息：[www.hnu.edu.cn,www.hnu.cn/湘ICP备05000239号] [hnu.cn 湘教QS3-200503-000481 hnu.edu.cn 湘教QS4-201312-010059]