A polyphyletic model for the origin of tRNAs has more support than a monophyletic model.

The monophyletic and polyphyletic origins of tRNAs are compared. The monophyletic hypothesis of the origin of the tRNA molecule would turn out to be strictly true only if a universal tRNA, possessing a number of anticodons with which to read primitive mRNAs, was present at the Last Universal Common Ancestor (LUCA) stage. On the other hand, whether a tRNA precursor less complex than it - for instance a hairpin - was present at the LUCA stage, then one should imperatively predict a phase of evolutionary convergence towards the cloverleaf structure typical of tRNAs, equivalent to a polyphyletic origin, also in presence of a real monophyletic origin. Thus, the monophyletic hypothesis is not able to eliminate the main trouble inherent in the evolutionary convergence, for which it is usually invoked. Arguments opposed to the existence of a universal tRNA and in favour of less complex precursors to tRNA, combined with the existence of two different classes of tRNAs, and, in particular, tyrosine tRNA being a class II tRNA in the bacterial domain but a class I tRNA in the eukaryal and archaeal domains, support a polyphyletic origin of tRNA molecules. It might have been the main path for the evolution of tRNAs, also because a polyphyletic origin of tRNA genes has been corroborated.