Revealing the conformation and properties of human genome, protein molecules and protein domain co-occurrence network

Abstract

Deoxyribonucleic acid, or DNA, encodes genetic instructions for the functionalities of organisms. For human beings, 23 pairs of chromosomes, containing DNA strands, form a globule structure in the nucleus. This chromosomal conformation influences the subsequent biological processes including transcription and translation by positioning sequentially remote genes spatially close. Chapter 2 reveals human chromosomal conformation and studies gene-gene interactions and “transcription factor binding site” interactions based on chromosomal spatial proximity. Proteins are the biological units that conduct biological functions. The three-dimensional structure of a protein molecule determines its particular functions. Chapters 3 and 4 discuss research in predicting protein tertiary structures. Algorithms that can predict residue-specific qualities of predicted structures were constructed and benchmarked. A knowledge database of soybean transcription factors is presented in Chapter 5, which contains predicted protein tertiary structures. Chapter 6 shows a computer system predicting protein functions using profile-sequence alignment, profile-profile alignment, and protein domain co-occurrence network. A biological process is usually performed by multiple proteins. Biological network provides a global perspective of studying lives, which usually considers the entire set of the same type of biological molecules of the target organism. Chapter 7 introduces a novel biological network, protein Domain Co-occurrence Network (DCN), and demonstrates that DCN has great potentials in inferring species phylogenies and predicting protein functions.