The present application discloses a method of identifying mutations in a target DNA sequence. The method involves:(a) hybridizing the target DNA sequence with a control DNA sequence wherein said control DNA sequence is the wild-type DNA sequence corresponding to the target DNA sequence to create a duplex;(b) treating the duplex to remove any spontaneous aldehydes;(c) reacting the duplex with a repair glycosylase to convert any mismatched sites in the duplex to reactive sites containing an aldehyde-containing abasic site;(d) reacting the duplex with a compound of the formula X-Z-Y, wherein X is a detectable moiety, Y is NHNH2, O-NH2 or NH2, and Z is a hydrocarbon, alkyhydroxy, alkylethoxy, alkylester, alkylether, alkylamide or alkylamine, wherein Z may be substituted or unsubstituted; and wherein Z may contain a cleavable group; for a sufficient time and under conditions to covalently bind to the reactive sites;(e) detecting the bound compound to identify sites of mismatches;(f) determining where the mismatch occurs; and(g) determining whether the mismatch is a mutation or polymorphisms.