That is one of the greatest questions of modern science. Given that even if a computer could sample every possible shape a protein could take it would still take an infinite amount of time given all the rotamers that are found in the literature. The short answer is the structure must fall down an "energy well" finding a global minimum of energy. The minimum energy is the native structure.
I also heard from my teacher in Biotechnical biochemistry (I think it translates to that in English) that they are working on finding segments or parts of proteins that can work as a base. So instead of trying tro predict every aminoacid they find many aminoacids which - if in that order - will always have the same structure and thus making it much easier to picture and understand the proteins.
@mangamagne this is exactly what knowledge based approaches use. They break up the sequence into fragments and see how these fragments usually fold. They then stochastically reassemble the fragments into a protein that meets the criteria of the energy function in question. Rosetta@home does this exact thing.
there are 8^n-1 possibilities. The proteins use the fold into the formation that uses the least energy, now to predict that.... you'd win a nobel prize probably if u could come up with a good way of how.
it depends on the protein but I believed its around 80 nano seconds.
Also folding@home is a program that allows your home computer to join in with thousands of others to help sim protein folding to fight disease and advance science.
Because of the huge amount of variables. If you change the position of 1 atom, you have to recalculated the forces acting within the entire protein molecule.
So to find the structure in which is a protein is completely stable (ie all forces are in an equilibrium) , you have to recalculated an enormous amount of times. At least, that's how I understood it ;)
That's only part of the problem, to minimize the energy of the structure you also need the right forces, and although we can estimate these quite well, it's still not a perfect solution. Another problem is that the folding of a protein is often helped by other proteins to make them fold the right way, so you would need to simulate those as well. Yet another problem are all the other molecules that surround the protein, mainly water, so you would need to simulate thousands (if not more) of these
molecules as well. All in all it takes huge supercomputers a lot of time to find the optimal conformations for a certain protein, and even then they often used a lot of known structural data of similar proteins to help find the solution. A simple solution for folding a certain protein correctly on your pc is probably still a couple of years (if not decades) away.
Well there's Folding@home project that makes use of your computer's/PS3's processing power joined with other thousands to simulate protein folding.
As far as I understand your unit would be running simulations for days on end to simulate small molecules and then sending the simulation results back to Stanford Uni.
What it does can be described as trial and error as it tries all possibilities until it works out the appropriate solution.
To simulate any process at atomic level, we need to calculate the interaction between each pair of atoms in the model system. Think about how many atom pair do you have if there are 100000 atoms in your system?!
However, two atoms that are separated far away do not interact effectively, so to save time, usually, only two atoms close enough to each other will be calculated. (That's called short range interation, and long range interaction would be ignored or estimated roughly) Still, you have a huge number of atom pairs to calculate.
DR. MALLERy"S CLASS IS THE BEST!
TheAzn5 4 months ago
@TheAzn5 Agreed! I'm sitting in it right now, and I'm riveted!
ramrune 4 months ago
Try an analogue computer.
ebutemetube 1 year ago
at what FrameRate does this video animates? 75Hz?
0731DENISE1996 1 year ago
@0731DENISE1996
If you can't even understand why that statement is wrong, how are you even aware of proteins?
snedie69er 1 year ago
Folding@Home ftw!
Marth8880 1 year ago
folding @ home
Misha070 1 year ago
Can anybody please tell me what is the background music?
Vitalik1Ru 1 year ago
Comment removed
mkoaah 1 year ago
My Protein is better than you Protein
knockdoun 2 years ago 2
I resent that, my protein hasnt been well, and it has lost its job.. shame on.. anyway your protein is gay.
robblac 2 years ago
flimsy looking.
Ibringthetruth1 2 years ago
Given an amino-acid sequence how does one predict the 3D shape that this protein will take upon folding?
rebreb06 2 years ago
That is one of the greatest questions of modern science. Given that even if a computer could sample every possible shape a protein could take it would still take an infinite amount of time given all the rotamers that are found in the literature. The short answer is the structure must fall down an "energy well" finding a global minimum of energy. The minimum energy is the native structure.
jrwkc 2 years ago
I also heard from my teacher in Biotechnical biochemistry (I think it translates to that in English) that they are working on finding segments or parts of proteins that can work as a base. So instead of trying tro predict every aminoacid they find many aminoacids which - if in that order - will always have the same structure and thus making it much easier to picture and understand the proteins.
mangamagne 2 years ago
@mangamagne this is exactly what knowledge based approaches use. They break up the sequence into fragments and see how these fragments usually fold. They then stochastically reassemble the fragments into a protein that meets the criteria of the energy function in question. Rosetta@home does this exact thing.
jrwkc 1 year ago
@rebreb06
there are 8^n-1 possibilities. The proteins use the fold into the formation that uses the least energy, now to predict that.... you'd win a nobel prize probably if u could come up with a good way of how.
Nibelhim 2 years ago
beautiful! How long does it take for a protein to fold?
sagun111 2 years ago
it depends on the protein but I believed its around 80 nano seconds.
Also folding@home is a program that allows your home computer to join in with thousands of others to help sim protein folding to fight disease and advance science.
beefalow2 2 years ago
depends how much washing was in the machine..
robblac 2 years ago
At 0:22 you can clearly see the face of Jesus.
ChrisArlingtonVA 3 years ago 2
crack me up!!
i'd say more a mixture of fergie and jesus!!
naomimamah 2 years ago
why does it take 2 weeks with a super computer to simulate 1 microsecond of protein folding?
Westonci 3 years ago 8
Because of the huge amount of variables. If you change the position of 1 atom, you have to recalculated the forces acting within the entire protein molecule.
So to find the structure in which is a protein is completely stable (ie all forces are in an equilibrium) , you have to recalculated an enormous amount of times. At least, that's how I understood it ;)
Twabi2 3 years ago 8
thanks
Westonci 3 years ago
That's only part of the problem, to minimize the energy of the structure you also need the right forces, and although we can estimate these quite well, it's still not a perfect solution. Another problem is that the folding of a protein is often helped by other proteins to make them fold the right way, so you would need to simulate those as well. Yet another problem are all the other molecules that surround the protein, mainly water, so you would need to simulate thousands (if not more) of these
finalart2005 3 years ago 2
molecules as well. All in all it takes huge supercomputers a lot of time to find the optimal conformations for a certain protein, and even then they often used a lot of known structural data of similar proteins to help find the solution. A simple solution for folding a certain protein correctly on your pc is probably still a couple of years (if not decades) away.
finalart2005 3 years ago 2
Well there's Folding@home project that makes use of your computer's/PS3's processing power joined with other thousands to simulate protein folding.
As far as I understand your unit would be running simulations for days on end to simulate small molecules and then sending the simulation results back to Stanford Uni.
What it does can be described as trial and error as it tries all possibilities until it works out the appropriate solution.
check this video EZ1XuOgknuE
ShiroKage009 2 years ago
2 weeks are nothing for microsecond simulation....for moderate size protein it can take months on supercomputer....
mkc1122 2 years ago
for what kind of calculation. de novo folding doesn't take that long.
jrwkc 2 years ago
@Westonci
Because there are too many atoms in the system we simulate: protein, water molecules...
xijunw 8 months ago
@Westonci
To simulate any process at atomic level, we need to calculate the interaction between each pair of atoms in the model system. Think about how many atom pair do you have if there are 100000 atoms in your system?!
xijunw 8 months ago
@Westonci
However, two atoms that are separated far away do not interact effectively, so to save time, usually, only two atoms close enough to each other will be calculated. (That's called short range interation, and long range interaction would be ignored or estimated roughly) Still, you have a huge number of atom pairs to calculate.
xijunw 8 months ago
Join team philippines - 2291 Support Folding@home.
tnx..
cyther39 3 years ago
what are these teams i see all over protein folding videos?
rollininoldstyle 2 years ago
join Folding4Ever - 110397 One of the best teams!
thesoftish 3 years ago
If anyone wants to join a team please feel free to join mine 104693
jicaley 4 years ago
GEEKS for a CURE Team ID #84451 Join Us GEEKS!
CoolPhotos 4 years ago
Beautiful video and even more beautiful music.
Any info on the music?
MickayelMkrtchian 4 years ago
I don't get it
ashleyy2919 4 years ago
Ummm, when do we get this version of F@H!?!? When we have 64 Core CPU's or 5 GHZ GPU's??
lookn2find 4 years ago
Look into running the SMP client in a VM environment... About a 25% increase over the normal SMP client.
Go to the high performance FAQ page on Sanford's website.
Arcygenical 3 years ago
can I get this video, I want to use it for an educational presentation
kingsh17 4 years ago
shall we dance?
voideye19 4 years ago
Very cool, but don't they fold starting at one end as they emerge from the ribosome?
NFormationFertilizer 4 years ago
cool!
sausagenmuff 4 years ago