Evaluation of the Persistence of Higher-Order Strand Symmetry in Genomic Sequences by Novel Word Symmetry Distance Analysis

Abstract

For the ubiquitous phenomenon of strand symmetry, it has been shown that it may persist for higher-order oligonucleotides. However, there is no consensus about to what extent (order of oligonucleotides or length of words) strand symmetry still persists. To determine the extent of strand symmetry in genomic sequences is critically important for the further understanding of the phenomenon. Based on previous studies, we have developed an algorithm for the novel word symmetry distance analysis. We applied it to evaluate the higher-order strand symmetry for 206 archaeal genomes and 2,659 bacterial genomes. Our results show that the new approach could provide a clear-cut criterion to determine the extent of strand symmetry for a group of genomes or individual genomes. According to the new measure, strand symmetry would tend to persist for up to 8-mers in archaeal genomes, and up to 9-mers in bacterial genomes. And the persistence may vary from 6- to 9-mers in individual genomes. Moreover, higher-order strand symmetry would tend to positively correlate with GC content and mononucleotide symmetry levels of genomic sequences. The variations of higher-order strand symmetry among genomes would indicate that strand symmetry itself may not be strictly relevant to biological functions, which would provide some insights into the origin and evolution of the phenomenon.

An Explanation of Exceptions from Chargaff’s Second Parity Rule/Strand Symmetry of DNA Molecules

“…Previous studies of human DNA sequences have shown that, in accordance with CSPR, the frequencies of direct and reverse complement oligonucleotides are approximately equal up to oligonucleotides with 6 nucleotides. With increases to higher order oligonucleotides to 7, 8 and 9 nucleotides, the differences between the frequencies of D and RC(D) become bigger, and after oligonucleotides with 10 nucleotides, the frequency differences change radically and sporadically and gradually break down, instead of breaking abruptly [ 35 ]. As seen, the answer to the question of why oligonucleotides with 10 or more nucleotides do not satisfy CSPR is shown in Table 2 .…”

“…For 11-nucleotide oligonucleotides, the CSPR analysis requires more than 6500 million nucleotides, i.e., the whole human genome would be too short. Such a gradual increase to longer DNA sequences for higher order oligonucleotides, because of the rapid growth of the number of oligonucleotide combinations, provides an explanation of why “the strand symmetry would break up gradually instead of breaking abruptly”, showing that strand symmetry persists for oligonucleotides of up to 9 nucleotides in the human genome for its oligonucleotide frequency pattern [ 35 , 55 ].…”

“…According to its meaning, this rule can also be called Chargaff nonlocal pairing. For 50 years, a conclusive explanation of CSPR has still been rather controversial [ 7 , 8 , 11 , 12 , 14 , 15 , 16 , 17 , 18 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 ].…”

“…CSPR for oligonucleotides cannot be explained solely by mononucleotides or lower-order oligonucleotide CSPR. Therefore, it has been suggested that it is likely that sequences were initially written at the oligonucleotide level under the evolutionary forces that required a parity at that level [ 19 , 28 , 35 ]. For the exact validity of CSPR, genes on both DNA strands must be exactly equal, and each sequence on the top strand (read 5′→3′) must correspond to the same gene sequence on the bottom strand (read 3′→5′).…”