Thread by @realscientists

This is one thing I didn't mention in my MD thread yesterday: solvent models 🧵

https://twitter.com/Cat_Arses/status/1381705179339235329

What's a solvent model? Simply, it's a way we can model processes that take place in solution (we'll use water as an example since it's the most common)

Proteins (large biological molecules) are usually surrounded by water so it's really important that we model these interactions correctly.

(We'll talk a bit more about proteins today)

Explicit model: this is probably the most intuitive, we model the water molecule with 3D coordinates (two hydrogens and an oxygen). This allows us to see what part of the water molecule is interacting with the protein as well as how the water molecules behave.

Implicit model: in this case we don't have any 3D coordinates to describe water molecules. Instead the effect they have on the solute (our protein) are modelled as a continuum. As long as the model reproduces the same properties as we'd expect then we're all good!

Another way to think of it: get in your car, stick your head out of the window, and drive down a road really fast - you'd feel the wind rush past your face (explicit model)

vs

buying a (very) large fan and sticking your face in front of it (implicit model)

In both scenarios you end up with the same result (wind in your face and a questionable hair style) but how you've got there is different.

(and yes I just wanted to put a dog GIF in the previous tweet)

Hybrid model: this is, as the name suggests, a mix of the two! You model some parts of your system explicitly and others implicitly.

Jumping back to explicit (water) models, there are lots of different types to choose from (each with their pros and cons). If you're interested, wikipedia has a good page on them here: en.wikipedia.org/wiki/Water_mod…

(a questionable twitter handle from the orignal post, but we're all friends here 🐱)