In Silico approach for differentiation of hypochromic microcytic anemia and dyslipidemia

• Main Author: Kandhro, Abdul Hafeez, 1959-
• Other Authors: Pornlada Nuchnoi
• Language: English
• Published: Mahidol University. Mahidol University Library and Knowledge Center 2023
• Subjects: Data mining; MicroRNAs; Dyslipidemias; Thalassemia
• Online Access: https://repository.li.mahidol.ac.th/handle/123456789/89801
• Institution: Mahidol University

Bioinformatics is a powerful research tools for facilitating disease diagnosis, treatment and prevention. In this study, the researcher systematically applied bioinformatics approach for solving common diseases/conditions in Asia. Unfortunately, the lack of special tests has hampered the differential and definitive diagnosis in poor resource setting areas. Several discrimination formulas have been used as simply screening tool to discriminate thalassemia traits from IDA. Therefore, twelve common existing discrimination formulas were applied in this study using Pakistan population as model. Six formulas (MI, EF, G&K, RDWI, R, and HHI) demonstrated the most reliability. Random Forest approach was used to improve cutoff values of the remaining six formulas. Furthermore, the researcher generated two new discrimination formulas with proposed cutoff values. The new formulas demonstrated high accuracies and the Youden's index. This indicated that the new proposed formulas and the six existing formulas with adjusted cutoff values discriminate IDA and thalassemia trait cases. We applied Text-mining approach was applied for constructing molecular network of miRNA related dyslipidemia. In order to comprehensively visualize the interaction of miRNAs and the role the related genes play in dyslipidemia, the researcher retrieved public databases information for microRNA and dyslipidemia associations, a total 227 associations including 148 microRNAs for constructing and visualizing interaction network by Cytoscape. The top 20 significant microRNAs used to derive predicted and validated gene targets by using CyTargetLinker. Predicted gene targets showed significant biological processes for cholesterol, lipid and fatty acids under the gene ontology analysis by BiNGO on Cytoscape. This study uncovered the complex network of miRNA in biological and pathological mechanism related to dyslipidemia. Finally, the impact of bioinformatics approach for laboratory application was demonstrated.