ASSESSING DISSIMILARITY OF GENES BY COMPARING THEIR RNASE-A MISMATCH CLEAVAGE PATTERNS

We propose a simple algorithm for estimating the number of nucleotide differences between a pair of RNA or DNA sequences through comparison of their RNAse A mismatch cleavage patterns. In the RNAse A mismatch c leavage technique two or more sample sequences are hybridized to the s ame RNA probe, the hybrids are partially digested with RNAse A, and th e digestion products are compared on an electrophoretic gel. Here we p rovide an algorithm for converting the numbers of unique and matching electrophoretic bands into an estimate of the number of nucleotide dif ferences between the sequences. Computer simulation indicates that the proposed method yields a robust estimate of the genetic distance desp ite stochastic errors and occasional violation of certain assumptions. Our study suggests that the method performs best when the distance be tween the sequences is <15 differences. When the sequences under analy sis are likely to have larger distances, we advise to substitute one l ong riboprobe with a set of shorter nonoverlapping probes. The new alg orithm is applied to infer the proximity of several strains of pseudor abies virus.