We determine conditions on network topology and preference correlations leading to extended clusters of people with similar interests.
Secondly, the RIASEC circumplex structure is evaluated upon the estimated latent preferences correlation matrix.
Therefore the strikingly high preference-occurrence correlations observed when GC content is close to 0.5 –that reach 99% in non-genic and intronic DNA (below)– may be due to the fact that when GC content is intermediate the primary-structural dynamics are favorable to the occurrence-preference correlation.
It shows that the evolutionary conservation of the occurrence-preference correlations is due to the conservation of both codon occurrences and motif preferences across species.
A important indication given by Figures 7 to 10 concerns the validity of the conclusion that the forces behind motifs preferences may shape vertebrate-DNA primary structure only or mainly when GC content is intermediate, simply because occurrence-preference correlations are highest when GC content is intermediate.
We find indeed that these preferences are conserved in vertebrates even more rigidly than codon occurrences and we show that the occurrence-preference correlations are stronger in intronic and non-genic DNA, with the R2s reaching 99% when GC content is ∼0.5.
Among the similarities between native and simulated coding-region occurrence-preference correlations are i) the clear lowering of some R2s relative to simulated non-coding ones, ii) the higher R2s for 4folds and 6folds than for the other three motif groups, and iii) the shift towards lower GC of the peak R2s of 4folds and 6folds.
Below, however, we will show that NBDM affects strongly the quality and quantity of the amino-acid replacements that point mutations generate so that it may not be a coincidence that occurrence-preference correlations are highest when GC content is intermediate, since that is the GC of most genes.
There was a clear preference for negative correlation, and the average correlation coefficients for the significantly correlated genes ranged from −0.5 to −0.3 (colorectal cancer series I and II, respectively).
This makes even more remarkable the match discussed above between the preference-occurrence correlations of groups of simulated and native sequences of similar GC content.
