Wavelets are mathematical tools used to decompose and represent another function described in the time domain, allowing the study of each component of the original function with a scale-compatible resolution. Thus, these transforms have been used to select conformations from molecular dynamics (MD) trajectories in systems of fundamental and technological interest. Recently, our research group has used wavelets to develop and validate a method, meant to select structures from MD trajectories, which we named OWSCA (optimal wavelet signal compression algorithm). Here, we moved forward on this project by demonstrating the efficacy of this method on the study of three different systems (non-flexible organic, flexible organic, and protein). For each system, 93 wavelets were investigated to verify which is the best one for a given organic system. The results show that the best wavelets were different for each system and, also, very close to the experimental values, with the wavelets db1, rbio 3.1, and bior1.1 being selected for the non-flexible, flexible organic, and protein systems, respectively. This reinforces our OWSCA as a very efficient and promising method for the selection of structures from MD trajectories of different classes of compounds. Our findings also point out that additional studies considering wavelet families are needed for defining the best wavelet for representing each system under study.