Mono- through hexanucleotide composition of the sense strand of yeast DNA: a Markov chain analysis

Here we compare several methods for predicting oligonucleotide frequencies in 392 kb of yeast DNA. As in previous work on E. coli, a relatively simple equation based on tetranucleotide frequencies can be used in predicting the frequencies of longer oligonucleotides. For example, the mean of observed/expected abundances of 4,096 hexamers was 1.00 with a sample standard deviation of .18. This simple predictor arises by considering each base on the sense strand of yeast to depend only on the three bases 5' to it (a 3rd order Markov chain) and is more accurate in estimating oligonucleotide frequencies than other statistical methods examined. This equation is useful in predicting restriction enzyme fragment sizes, selecting restriction enzymes that cut preferentially in coding vs noncoding regions, and in constructing detailed physical maps of whole genomes. When ranked highest to lowest abundance, the observed frequencies of oligomers of a given length (up to 6 bases) are closely tracked by the predicted abundances of a 3rd or 4th order Markov chain. These ordered abundance curves have a power curve shape with a broad linear range with a sharp break at the top end of the curve. There is also a strong disparity between the most and least abundant oligomer with for example a 79-fold variation between the most and least abundant hexamer. The curves reveal a strong dependence of oligomer frequencies on base composition. Unlike E. Coli, there is no sharp downturn at the low end of the curves and hence, no class of oligomers rare relative to other oligomers of the same length.