MODELING PROPERTIES OF BIOCHEMICAL-COMPOUNDS WITH CONNECTIVITY TERMS

The descriptive and utility power of linear combinations of connectivi ty terms (LCCT) derived by a trial-and-error procedure from a medium-s ized set of 8 connectivity indices: {chi} = {D, D-v, (0) chi, (0) chi( v), (1) chi, (1) chi(v), chi(t), chi(t)(v)} or from a subset of it has been tested on properties of heterogeneous classes of biochemical com pounds centered on the homogeneous class of natural L-amino acids. To choose the appropriate combination of indices the forward selection an d the complete combinatorial technique have been used, whenever more t han a single term was necessary for the description. The forward selec tion technique searches only a subspace of the complete combinatorial space, but nevertheless has many advantages among which to be a good t ool for an elementary and direct test for newly defined indices. The m odeling has been followed centering the attention not only on the pred ictive power of the proposed linear equations but also on their utilit y. The modeling of the solubility of the entire heterogeneous class of n = 43 amino acids, purines and pyrimidines could satisfactorily be a chieved with a set of supraconnectivity terms based on the chi(t)(v) i ndex mainly. The unfrozen water content of a mixed class of inorganic salts and natural amino acids has satisfactorily been modeled with two connectivity terms and the modeling shows a remarkable utility. The u tility of the given LCCT can nevertheless be enhanced, especially when the modeling requires 2 or more terms, with the introduction of the c orresponding orthogonal indices, as can be seen for S(AA + PP) and UWC . Further, the delta cardinal number is used as starting point for the definition of a supravalence index a to be used for a topological cod ification of the genetic code and the amino acids in proteins. In fact , the notion of supravalence can be extended to the triplet code words to generate the different families and subfamilies of the genetic cod e and to visualize the connections of amino acids in proteins. Three p roperties of the DNA-RNA bases (U, T, A, G and C), the singlet excitat ion energies Delta E-1 and Delta E-2, and the molar absorption coeffic ient epsilon(260) have been simulated with a single connectivity term chosen from the same medium-sized set of 8 molecular connectivity indi ces.