I am having many singleton Species (species representing only one DNA sequence). These singletons lack resolution potential as they are single, when applied machine learning classifiers singletons are not considered as there is no reference for it... Even using other relevant methods, we cannot resolve singleton species confidently.

Please suggest any method/program that could simulate or generate reference sequences by taking those singletone species into consideration. Further this reference sequence could be used to resolve singleton species in the multiple species sequence dataset.