In my mind, starting from a reliable tutorial is usually the most efficient way to learn new technology. After going through the whole process from setting-up to the final results, you may have a better insight of the method and may have some ideas for your own simulation.

https://plumed.github.io/doc-v2.3/user-doc/html/belfast-6.html

Stick to using software packages as described above especially the Belfast tutorial set, but in terms of journals these should be useful to help tune the collective variables. You'll find that metadynamics for proteins is very difficult. There are numerous collective variables to consider so picking the right one is difficult. Often, PLUMED can cause gromacs to throw out LINCS errors as the protein is pushed into very unlikely and troublesome conformations.

Article Enhancing Important Fluctuations: Rare Events and Metadynami...

Article Metadynamics

Article Enhanced Sampling Techniques in Molecular Dynamics Simulatio...

Article Practical Aspects of Free-Energy Calculations: A Review

Article Using Collective Variables To Drive Molecular Dynamics Simulations

http://www.pnas.org/content/99/20/12562

Thank you very much