I doubt it's in the class of problems that would benefit from quantum computing.
The problem is that it's a subclass of the N-Body problem
http://en.wikipedia.org/wiki/N-body_problem
Bigger/faster computers help a little but it costs O(N^2) where N is the number of particles you're simulating meaning that to simulate a slightly bigger protein or pair of proteins you'd need a vastly more powerful computer.
Add any slight errors in our models for subatomic particles.
Add in simplifications depending on how low a level you want to simulate it.(Atoms? electrons/protons/neutrons?)
I've seen high quality simulations of protein interactions where they simulate the forces between each atom but they take a long time and a huge amount of computational power.
Unfortunately predicting how one 3d protein will bind to another protein is very hard.
The problem is that it's a subclass of the N-Body problem
http://en.wikipedia.org/wiki/N-body_problem
Bigger/faster computers help a little but it costs O(N^2) where N is the number of particles you're simulating meaning that to simulate a slightly bigger protein or pair of proteins you'd need a vastly more powerful computer.
Add any slight errors in our models for subatomic particles.
Add in simplifications depending on how low a level you want to simulate it.(Atoms? electrons/protons/neutrons?)
I've seen high quality simulations of protein interactions where they simulate the forces between each atom but they take a long time and a huge amount of computational power.
Unfortunately predicting how one 3d protein will bind to another protein is very hard.
You may have been given a harder problem than your supervisor realized they were giving.
I can't think of any easy way to approach your problem.
Warning, semi-random musings, not recommended to try:
Possibly examine a selection of binding sites, pick out the common features important to the binding, convert each of them back to their sequences, compare to pick out the common parts that have high information content, come up with some kind of set of weighted masks then search for genes that contain sequences matching them. Though that's just off the top of my head so it's very likely to not work, have insurmountable problems, find an insane number of false positives, be a huge pain to even try and would still be computationally expensive but in a "I need a big computer" way rather than an "I need a computer the size of the planet" way.
On Thu, Aug 7, 2014 at 4:04 PM, Cathal Garvey <cathalgarvey@cathalgarvey.me> wrote:
It's not a matter of bioinformatic advancement, it's a matter of
computational advancement. That sort of prediction assumes you can
physically compute things like protein folding, binding affinities and
such on a conventional computer, which is not presently possible to a
useful extent.
You can model proteins with computers, but they're not very good at it
and it takes ages to do. You could then try to predict binding
affinities of ligands to the proteins, but doing so would involve
trial-and-error computation and take a long time and be very error prone.
I'm not a computer scientist, but my intuition suggests that this sort
of problem might benefit from quantum computers. So, maybe in our
lifetime we'll have protein structure solvers and protein ligand
predictors that don't suck. But right now empiricism is the only
reliable way.
T: @onetruecathal, @IndieBBDNA
On 07/08/14 14:26, Mega [Andreas Stuermer] wrote:
> Hm ok. But bioinformatics is not advanced enough to make prediciton on
> ligand binding? I read that these methods are all empirical from previous
> studies
>
--
P: +353876363185
W: http://indiebiotech.com
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diybio@googlegroups.com. To unsubscribe from this group, send email to diybio+unsubscribe@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+unsubscribe@googlegroups.com.
To post to this group, send email to diybio@googlegroups.com.
Visit this group at http://groups.google.com/group/diybio.
To view this discussion on the web visit https://groups.google.com/d/msgid/diybio/CAP0RkBC1pGSQzO1sVAy5Wk8wH6ERhkXkud-qH2Fadkwo%3D%2BiJTQ%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.
0 comments:
Post a Comment