My current research is based on measuring a melting temperature of PCR-products of one gene. We are tryng to find all polymorphisms among healthy population by sequencing only those PCR-products that have different melting curve from wildtype variants (to lower expenses from not to sequence all PCR-prods.). Often, it is very difficult to distinguish melting curve of polymorphic from wildtype just by looking at it! Is there a chance to build a decision tree that would say if one PCR-product is polymorphic or wildtype based on data collected from measuring a melting temperature?! Or should I use any another method to do so? Please, help!