To give an answer to this kind of questions is always based on what you wanna calculate from your simulations. The best way is to make a literature review and go through available papers and end up with a list of possible softwares that can be used for your purpose.
Answers people are gonna give you here would be confusing for you, because each researcher will probably recommend you the software that they have had experience. And no one knows what property you wanna get from your MD simulations. Your questions is not even self-explanatory as well.
Protein folding should be approached by experiments in solution, often by means of biophysical techniques. The computer approach should be used to validate a model or to gain some structural information to drive the experiments in solution. Sometimes, when an experimental plan is difficult to be performed in solution (for example, membrane proteins), the use of the computer (molecular dynamics) may be useful. However, the computer should not be the first choice to work on folding for not skilled people, but, since the protein family is made by globular as well as intrinsically disordered proteins, in this last case the computer approach could be a good choice. 3D Modeling software and GROMACS together with PyMol are important tools. The software that make network analysis using amino acid residues should not be underestimated in studies of protein folding. In this case Cytoscape is a useful tool of visualization.