Applications of different computational approaches to study proteins in biological systems will be discussed in this talk.We will first show how molecular dynamic(MD)simulations with QM/MM potentials can be applied to understand the catalytic mechanisms of individual enzymes or enzyme complexes as well as their specificity.It is demonstrated that such computational approaches can serve as pow-erful tools for understanding details of enzyme catalysis.As we know,determination of complete genome sequences from a number of organisms have offered an unprece-dented opportunity for biological research and transformed biology into a discipline that depends significantly on how to interpret large-scale data sets.By selecting representative proteomes from three domains of life,two giant DNA viruses,and col-lective gene sets from viruses and organelles including mitochondria,chloroplast and plasmids,we will demonstrate that systematical analyses of the interplay between protein length(L,i.e.,the amino acid sequence length)and protein disorder(D,i.e.,percentage of residues in a so-called intrinsically disordered state)may allow us to construct a two-dimensional LD-space for describing the proteomes or gene sets.It is found that the gene distributions in this LD-space may serve as an architectural fingerprint shaped by the evolutionary processes.