Thanks for a great video. Just one query. Isn't Coulombic interaction inversely proportional to the square of the distance of separation? You haven't mentioned power of 2.
17:23 Although slightly better (cuts the number of calculations by two), it's still of the QUADRATIC order. Is there any better way of calculating interactions between particles than going through each pair one by one? I know that this could be done for collision detections, by subdividing space in a recursive way (e.g. octrees or K-D trees) and excluding all those pairs that are too far away to be able to collide, but I don't see how the same trick could be applied to forces, since they can still act even at large distances :/ (sure, we can neglect all those interactions for atoms that are too far away and the force of interaction is negligible), but I don't thing that would make the cut. Are there any better ideas?
That's about it really. You make an approximation and only calculate interactions upto a certain cutoff. For the vDw terms this gives very small errors as the value decays fast with respect to distance. For electrostatics, you divide the interactions into two parts, where the short range interactions are calculated directly from the equation in the video, but the long range interactions are determined from the Fourier transform of real space. Check out the methods titled, "Ewald summation" or "Particle mesh ewald".
If anyone is interested in protein modelling and molecular dynamic simulations , you may refer to my channel I teach the basics of VMD program which is used widely in molecular dynamic simulations and protein structure analysis th-cam.com/play/PLhYF9QNr23IaG-aFub24_T8XnIbJb7fNU.html
Just happened upon this video by chance. I had to move on early in the presentation when you indicated that you must know position and momentum simultaneously at the outset. Forgive me if i missed some bigger concept. Will take my chances though.
Fanstastic explanation, but I think there is a mistake (or I did not get it probably). The first force you calculated was a derivative of energy with respect to distance but the starting structure at t=0 has the atoms at a constant equilibrium distance (for instance an energy minimized structure). So without first taking a step (t = 0 you said), the force will be zero. If it is otherwise, however, it means you have taken arbitrary step to compute the force. Please let me know what you think. Thank you
Very nice explanation. However, I have always wondered why initial velocities are required when setting up a MD simulation? Does the energy of the system (and the resulting forces on the atoms) also depend on the initial velocities?
xabier arias yes in reality initially in fluids every particles is moving with some velocity because only at absolute zero temperature atoms or molecules won’t have any moment... rest of the cases they have finite momentum
I think in depends. For example if you're simulating just one protein molecule then every atom in the molecule may be started at velocity zero, because at the beginning of the simulation you don't have any idea of the velocity of each atom
Thanks a lot for simplistic explanation of the concept. Now i have some queries. between 28:33 to 29:06 you mentioned different positions of the atoms. But where in the direction of motion is changing? which parameter is reflecting about that direction? Though its denoted as vector, i am still not clear about how this is being taken care and in what way? Wish to have a reply from you or anyone, please. Happy new year.
The best explanation I could get for understanding what's happening during the molecular dynamics simulations
I have taken a course on Basic Simulations and this really helped me , your explanation was far more better than at my university. Helped me alot.
If I may, *far better. NOT far more better.
This was an amazing video. thank you! I just wanted to make sure that everything I know is right or not.
After going through a lot of convoluted material on this subject, this video is a breath of fresh air.
This talk is simply brilliant, you have saved my life, thank you 🍀
sometimes we know how to run a simulation but don't know what we are exactly calculating. this simple video is great!
This introduction is concise and clear!!! Thumb up !!!
Perfect sir ! The most useful video on MD for noobies, so far :)
This is pure Gold ! . Very informative, Thanks a lot
This was so well explained. I was having a hard time understanding it by reading only. Thank you.
Hmmmm
Could you make a video explaining (Mesh) Ewald Summation, which seems to play a big role in MD?
Was very helpful to get an insight into what molecular dynamics is. Eager to see more such videos. :)
I really enjoyed your explanation of the MD cycle around min 20. Thank you.
Very useful video, teaches the theory behind the simulation with so much ease
Nice and simple presentation. Thanks for sharing your knowledge with us..wish there were videos on deeper concepts..
Many thanks
Clear and comprehensible introduction!
Thank you so much for this lucid and concise introduction!
This was actually a very lucid way of explaining this topic! Loved it!
Thank you very much for the course, which is very informative for me as a beginner.
great explanation! God bless you, Mat!
Wonderfully explained, thanks so much for making this for us!
Its very straight forward and easy understanding! Thanks for your teaching!
A quick intro to MD. Thank you.
Good video. Tahnks. got some basic idea of MD simulation. and the direction of accelaration is decided by the Force vector. right?
Thanks for a great video. Just one query. Isn't Coulombic interaction inversely proportional to the square of the distance of separation? You haven't mentioned power of 2.
Santanu Roy Thanks for watching! Coulombic energy decreases as 1/r, and correspond force decreases as 1/r2.
@@MBauchy Oh yes, sure. How stupid of me.
Wonderful lecture, Thank you very much!!!
17:23 Although slightly better (cuts the number of calculations by two), it's still of the QUADRATIC order. Is there any better way of calculating interactions between particles than going through each pair one by one? I know that this could be done for collision detections, by subdividing space in a recursive way (e.g. octrees or K-D trees) and excluding all those pairs that are too far away to be able to collide, but I don't see how the same trick could be applied to forces, since they can still act even at large distances :/ (sure, we can neglect all those interactions for atoms that are too far away and the force of interaction is negligible), but I don't thing that would make the cut. Are there any better ideas?
That's about it really. You make an approximation and only calculate interactions upto a certain cutoff. For the vDw terms this gives very small errors as the value decays fast with respect to distance. For electrostatics, you divide the interactions into two parts, where the short range interactions are calculated directly from the equation in the video, but the long range interactions are determined from the Fourier transform of real space. Check out the methods titled, "Ewald summation" or "Particle mesh ewald".
Nice presentation on molecular dynamics simulations sir
Great videos! Thank a lot. But sometimes hard to read your handwriting.
yep
C'était très clair, tu gères !
Merci :)
Thanks a lot for this awesome video! perfectly explained!
thank you it was really useful. I would like to learn more about MD
How the hell he was able to deliver all of this and I mean simply its dope.
Thank you. Helpful to understand the basics
Many thanks, thanks for watching!
If anyone is interested in protein modelling and molecular dynamic simulations , you may refer to my channel I teach the basics of VMD program which is used widely in molecular dynamic simulations and protein structure analysis
th-cam.com/play/PLhYF9QNr23IaG-aFub24_T8XnIbJb7fNU.html
Got sparked !
Want more on that topic.
Just happened upon this video by chance. I had to move on early in the presentation when you indicated that you must know position and momentum simultaneously at the outset. Forgive me if i missed some bigger concept. Will take my chances though.
Thanks for this explanation. really helped
Would I need a solid understanding of quantum mechanics to excel in this area?
life saver...thanks a lot...plzz keep more videos coming
Thanks!
Fanstastic explanation, but I think there is a mistake (or I did not get it probably). The first force you calculated was a derivative of energy with respect to distance but the starting structure at t=0 has the atoms at a constant equilibrium distance (for instance an energy minimized structure). So without first taking a step (t = 0 you said), the force will be zero. If it is otherwise, however, it means you have taken arbitrary step to compute the force. Please let me know what you think. Thank you
Thank you very much for such a simple explanation....
Many thanks
Very nice explanation. However, I have always wondered why initial velocities are required when setting up a MD simulation? Does the energy of the system (and the resulting forces on the atoms) also depend on the initial velocities?
xabier arias yes in reality initially in fluids every particles is moving with some velocity because only at absolute zero temperature atoms or molecules won’t have any moment... rest of the cases they have finite momentum
@@balaram4967 True! thanks for the reply
I think in depends. For example if you're simulating just one protein molecule then every atom in the molecule may be started at velocity zero, because at the beginning of the simulation you don't have any idea of the velocity of each atom
Thanks for this special presentation.
Excellent presentation. Thanks a lot!
Very helpful information..Thanks for your effort..
Thank you very much, very helpful and useful.
Many thanks!
Thanks a lot for simplistic explanation of the concept.
Now i have some queries.
between 28:33 to 29:06 you mentioned different positions of the atoms. But where in the direction of motion is changing? which parameter is reflecting about that direction? Though its denoted as vector, i am still not clear about how this is being taken care and in what way? Wish to have a reply from you or anyone, please. Happy new year.
Thank you very much! Very useful explanation!
wow, great video, hope you keep uploading videos. Thanks a lot!
Thanks!
Perfect , waiting more videos like this ..please go on
Thanks for watching!
Nice and informative video.Thanks Sir.
Love it - thank you very much
Thank you for informative lecture.
thank you.... please upload tutorial videos of molecular dynamics simulations in LAMMPS
Check out this video: th-cam.com/video/QdygibMW1AI/w-d-xo.html
Thank you its usefull and clear explanation
Good video presentation.
Thanks alot. great video for beginners
Great video! Thank you so so much!
plzz upload a video discribing the statistical mechanics needed for molecular dynamics
Please subscribe to the channel to receive the future videos from our group.
Thanks so much, this was very helpful.
Thanks for watching!
❤ great explanation!!
Very well explained
Great video!
great explanation!!
Amazing video sir
Very informative session.
Nice session
THANK YOU FOR THIS!!!
Hi there. keep it up a great work.
Thanks for watching!
@@MBauchy Hi! may i know are you researcher/scholar or molecular modeler?
Tq Dr . :)
wonderful
Thank you for information
Sir, I like to do some research collaboration.
thank you very much
its an amazing video.
you are great thanks
Thank you
Fantástico!!!
just amazing thank's
Thank you sir!!!
It's great video
Actually very useful, big likkkkkke
MD 15:14
Interesting.
Good
Damn, too many ads
🙏...
Your ones are [A]mazing
❤❤❤❤❤❤❤❤❤❤❤❤❤❤
wow!!!!!!
Thank you so much for this lucid and concise introduction!
Thank you very much! Very useful explanation!
thank you very much, the explanation was great
Thank you for informative lecture.