There is no right answer, of course. Choosing a first language will depend on what kind of projects you want to work on, who you want to work for, or how easy you want it to be.

The first to you need to study is algorithm design. Programming language does not matter.

Python has easy use, and provide fast plotting libraries & has this useful frameworks: http://www.introductiontographene.org/python-packages-for-graphene-based-materials-and-beyond/

I am by no means an expert in this field but i think you should consider C++ to do it. Molecular dynamics simulations are computationally expensive and have high potential for scaling up i.e parallel programming. So it is worth your while to implement it in C++. As a matter of fact, in our university, we have codes related to molecular dynamics and lattice boltzmann methods in fluids, that are developed completely in C++ (sry for the anecdotal evidence!).

Python (or Matlab, if you can afford it!) are generally good for developing code initially. But later if when speed becomes an issue (which it definitely will be, once the problem size becomes large enough), then you can shift to C++. Also you can use python wrappers over your C++ code. This way, you can use other python libraries, especially visualization tools. By the way, if you decide to use python, definitely vectorize your code.

tldr: C++ is better for performance and python for fast development. You can learn them both and combine.

If you are not an expert programmer then Python is the best choice. It is easiest to learn. If you lack performance, then compile your working Python program with Cython.

(I teach Python, C, C++, Java and several other less common languages at a University)

I do everything in C because I've been using it for nearly 20 years. The "best" one is whichever one you can use to get the job done correctly.

That said, python is wonderful and I wish I was better with it.