Objective Distances for Transcriptome Analyses According to the Thermodynamic Model

Data analysis and interpretation require intellectual frameworks, and to share the results in a community, scientific bases are preferred for the framework. A variety of indexes have been used for interpreting distances between transcriptome data, however, such indexes always lack necessity since they have been derived for non-biological purposes and their appropriateness have not been verified for transcriptome analyses. Here introduced are a definition of distance and two indexes that are derived from thermodynamic model of transcriptome formation [3, 4], which describes transcriptome by way of interactions between protein factors and nucleic acids. For example, a human cell may have 1,000 species of protein factors, and its transcriptome can be described by the same numbers of dimensions of Gibbs free energies and Arrhenius activation energies. Consequently, distance between any two expression levels can be found as Euclidean distance calculated with the energies. However, unfortunately, such data treatment is presently impossible, since the constant relationship between a protein factor and a gene is not available yet.