Naslov
Novel Graphical Representation of Proteins Based on the Virtual Genetic Code

Autori
Randić, Milan ; Vračko, Marjan ; Plavšić, Dejan

Izvornik
International journal of chemical modeling (1941-3955) 1 (2008), 3-4; 305-324

Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni

Ključne riječi
Protein ; Virtual genetic code ; Graphical representation

Sažetak
We consider construction of graphical representations for proteins, which is based on use of the virtual genetic code. The virtual genetic converts protein structure expressed through amino acids into the sequence of nucleotides adenine, cytosine, guanine and uracil, (tyrosine in case of DNA) A, C, G and U (T), respectively, bypasses the degeneracy of the natural genetic code by assigning to each amino acid a single codon. We show that this mathematical simplification preserves and reflect correctly inherent similarities and dissimilarities among proteins as illustrated on protein ND6 for human and gorilla. Moreover, it is shown that the selection of a particular virtual genetic code is not the critical factor, and that alternative construction of 1:1 correspondence between amino acids and nucleotides would produce the same results. Our conclusion is based first on a direct comparison of the difference of the spectral representations of the two ND6 proteins (having 174 bases) based on amino acids (involving 20 distinct numerical values for the 20 amino acids) and based on the difference of the spectral representations of the same two proteins based on four nucleotide (having length 522 bases) involving four distinct numerical values for the four basis. Then we considered a compact 2-D map representation of the two proteins based on the virtual code (that is four the bases A, C. G and U) and the “worm”-type spiral as a template, in which the four bases A, C. G and U are coded by binary codes: 00, 01, 10 and 11, respectively, and again a complete agreement is found in the difference plots between the ND6 proteins of human and gorilla based on the binary codes and those based on the protein amino acid structure. Finally, we constructed a six-digits binary code for the twenty amino acids, which transformed the two proteins into corresponding binary sequences (of length 1044) and demonstrated that all these different protein representations always produced the same difference plots for the ND6 proteins of human and gorilla. The novel graphical representation of proteins introduced in this contribution, and used for comparison of ND6 proteins of human and gorilla, is based on the graphical representation of the virtual genetic code as a portion of the 8x8 table of codons. This results demonstrated beyond doubt that use of virtual genetic code, which bypasses the synonymy of the codons of the natural genetic code, is not only practical but is legitimate from the scientific point of view of bioinformatics, if not biology.

Izvorni jezik
Engleski

Znanstvena područja
Kemija

POVEZANOST RADA