Also known as "genomic signature" or "codon signature", this reflects the non-randomness in the arrangement of neighboring nucleotides. "Neighboring" here does not imply "adjacent", but merely conveys that the nucleotides are within several bases of one another along a linear strand. Exactly what constitutes "short" depends on the nature of the sequence and the phenomena under consideration. Generally, we are talking about a range of less than 10 bases.

A new term that we coined to refer to the non-randomness in the arrangement of nucleotides at a distance of >20 but <10,000 nucleotides. Exactly what constitutes "mid-range" depends on the nature of the sequence and the phenomena under consideration. Generally, we are talking about a range of 30-1000 bases.

This term refers to the presence of mid-range inhomogeneity throughout the genome (that is, not limited to coding regions, untranslated regions, or even to genes). Note that it is advisable to work with only sequences that can be considered "uniform" from a genomic perspective. By that we mean that one would likely want to work with only 5'UTRs or only introns or only intergenic regions from a particular isochore type, since this would avoid mingling sequences with very different nucleotide compositions.