Hi Tini Chu

Total depend on simulation software. First of will be understand about fundamental of software and how to execute the required parameters (like RMSD, RMSF, rGYR etc). So very carefully step by step follow tutorials and prefers some research articles.

First describe your specific area of research like DFT, MD, etc. Then you could be benefited with the people's comments.

Tini Chu, Here are some set of lectures given by me in my Youtube channel. Those might be helpful for you. Please have a look.

1) Theory of Molecular dynamics : https://youtu.be/UVCqsm4iWiw

2) Write your first own MD simulation : https://youtu.be/RPVfDXRRorU

3) Apply a Thermostat : https://youtu.be/X3K5A96-U4E

There are also other videos which you may like. Please subscribe to my channel to support me. Thank you.

Waseem Ahmad Ansari Hello, thank you for answering, I will try to follow more articles

Shashi Kumar Hello, I only do MD studies(I even just googled dft and get back...)

Souvik Sadhukhan Hi, I have just subscribed! I will have a look later. Look forward to the upcoming videos

By the way, I concluded that I should follow more articles and see how people do research on targets of their interests (i.e. knowing the structure and properties, thus to set the parameters). Is that true? Or is there anything I lost or mistook?

Thank you in advance :)

Tini Chu Yes that is true. There exists a very nice book on Molecular Dynamics by Frenkel Smidth. Have a look on that.

Dear Tini Chu

To properly get started you need a solid theoretical background of thermodynamics, statistical thermodynamics and molecular dynamics in general. Many good books have been written over the years, such as

Computer simulation of liquids (Allen)

Understanding molecular simulations (Frenkel)

Molecular modelling (Leach)

Thermodynamics and statistical mechanics (Greiner)

Once you have the theoretical tools, you need the practical ones. This mainly means finding a software which is able to run simulations of your system of interest, depending also on the hardware available. GROMACS, CHARMM, LAMPS are common examples of MD software. You have to decide what to use and install it if not already available to you.

Next, and arguably most important step, is setting up your starting systems, which roughly reduces to choosing a force field. What system are you interested in? What molecules? What timescales? Is the force field implemented in the MD software you are using?

At this point you will be able to start running simulations about your system of interest.

There is no general answer to these questions. Some software is better than another in certain aspects, or supports a force field that you need while others don't. Similarly, some force fields work better for some properties/molecules/systems than others. Maybe you are looking to be fast, and therefore are looking to coarse grained or united atoms systems.

A 'must do' is looking in the given field and understand what has been done and what is the common procedure to do it. If you are interested into binding properties of some ligand, then read relevant literature about it, maybe something has already been done with your ligand of interest. What software are they using? What force field? What properties do they measure? Once you have an idea, install and get experience with that software. For example, for GROMACS do some of these tutorials (first one always plus other relevant to you)

http://www.mdtutorials.com/gmx/

I use GROMACS as an example because that's what I work with, but CHARMM, LAMPS etc are available and you will for sure find many tutorials online. Once you are familiar with the software, find some fundamental work in the literature that is achievable and try to reproduce it, so that you get experience with a system similar to yours and with the sampling procedures you also will need.

Hope this helps,

Nicola

Numerical simulation is a collection of tools that can be used to elicit information about a molecular system that can be difficult to obtain experimentally or to provide atomistic interpretations of experimental results. Each tool in the collection has its purposes, weaknesses and strengths. The most important thing for you to decide is what is the scientific question that you are trying to answer. Then, select a tool appropriate to that question.

There are a host of different software packages and force fields available, most can be obtained without charge. I used the CHARMM force field and NAMD/VMD in my researches but that was mostly happenstance. My mentor was part of the CHARMM development team and I found the NAMD software to be the first that really implemented parallelism. As a result, I can provide some handwaving reasons why CHARMM is superior to Amber but the differences are actually quite small.

If there is someone at your institution who is already conducting simulations, it would likely be best to just adopt their choices and practices. This strategy will give you built-in technical support locally. If not, then flip a coin and choose one. No choice will be wrong. Install the software and start performing the tutorials. There is a huge learning curve that cannot be avoided.

Again, it is of primary importance that you define the scientific question before embarking on conducting a series of simulations. It's very easy to burn (thousands of) hours of CPU time without providing any clarity.

Sorry for the late reply, I was having midterm week.

Thank you all so much for answering, your suggestions are really helpful! I will try to get familiar with this field as suggested. There's a bunch of work I should do before carrying out my own research haha...