Similarity and dissimilarity in correlations of genomic DNA

We analyze auto-correlations of human chromosomes 1-22 and rice chromosomes 1-12 for seven binary mapping rules and find that the correlation patterns are different for different rules but almost identical for all of the chromosomes, despite their varying lengths and GC contents. We propose a simple stochastic process for modeling these correlations, and we find that the proposed process can reproduce, quantitatively and qualitatively, the correlation patterns found in the genomes of human and rice.