The frequency of oligonucleotides in mammalian genic regions

The large body of nucleic acid sequence data now available offersa unique opportunity for the characterization of individualoligonucleotides which may be specific to sequence functionaldomains. We have prepared algorithms for the study of the frequencydistribution of all oligonucleotides of length 2–6 inDNA sequences. We have implemented them in the study of 634mammalian DNA sequences spanning 1.782 Mb, and have obtainedthe distribution of the ratio between the observed frequencyof oligonucleotides and their expected frequency based on independentnucleotide probabilities. We then studied the distribution ofoligonucleotides (or k-tuples) of each length in a subset of129 complete mammalian genes spanning 0.607 Mb. Eight distinctgenomic regions, namely 5'-non-transcribed, first exon, firstintron, intermediate exons, intermediate introns, last intron,last exon and 3non-transcribed, were considered. We observedthat some oligonucleotides show a statistical behaviour anda regional distribution similar to that of known signal sequences.Moreover the frequency distribution of oligonucleotides of length5 and 6 tends to become bimodal, indicating the existence ofa population of very frequent oligonucleotides. Received on June 21, 1988; accepted on October 14, 1988