Hello there,
I am going to model a simple reaction between Fe from heme of human erythrocyte catalase and hydrogen peroxide using QM methods. Before that I need to have valuable structure of active center. What is the best approach to obtain it?
I have found in some research papers that one way consist of exporting structures within 15-20A from the point of the interest (so herein it would heme prosthetic group and some of the surrounding amino acids), adding hydrogen atoms and then running quantum chemistry calculations with frozen labile atoms of the cut off amino acids in a required solvent (let's say PCM water).
On the other hand, some studies insist on using ONIOM methodology, however, in my case I don't find it necessary – it's just a simple reaction. Iron heme seems to be linked with a nearby tyrosine, but nothing more. Yet, I'm not sure whether interactions of further parts of porphyrin with amino acids would affect Fe activity so greatly to include them.
What would you recommend me? I must aware you, I am able to use just freeware software - most likely VMD.
You need the ONIOM calculation. PCM is just an approximation and you never know what will happen if the truncation (according to your distance 15-20A criterion) is sufficient. Further, QM/MM results show significant dependence on the size of QM region. ONIOM would be the best and a fail-safe method to apply.
You need the ONIOM calculation. PCM is just an approximation and you never know what will happen if the truncation (according to your distance 15-20A criterion) is sufficient. Further, QM/MM results show significant dependence on the size of QM region. ONIOM would be the best and a fail-safe method to apply.
Dear Maciej,
I second/support the advice given by Zhaoxi. However, I would have several additional pieces of advice.
1) Physics is built on the assumption of existence of an isolated system.
2) This physical assumption is ill suited to almost any biological system.
3) As a student you have the duty to find out what is this proper boundary limit for you to obtain reasonable results.
4) Recent advances in protein folding and in allostery clearly suggest that such a boundary must include a lot of atoms. Both theories advanced by including a global free energy transformations. What it practically means is that inclusion of a single water molecule on the periphery of the protein under some critical circumstances may completely change the outcome. Read a bit about allostery in bivalve hemoglobins.
5) The general conclusions are: (A) Use as large a system as practicable, (B) use as small QM system in INIOM as possible to obtain repeatable results, (C) There is no real theory in theoretical physics applied to biology, so use as many computer experimentation (separate runs) as possible, if applicable use replica tricks, (D) If practicable use density-functional methods, (E) You must establish what solvent molecules, what pH and what other factors contribute to your results. (F) The most important piece of advice is an appropriate selection of the protein from the PDB structures. The protein must be of high resolution (best, higher than 1.2Å) very high quality (better than 0.13 in Rfactor), very well modeled density with residual features less than a fraction of electron per Å3 , so all the necessary disorder and solvent structure is included. Proteins work similarly to mechanical devices so expect sometimes large movements even for simple chemical reactions. You must learn about proper preparation of the protein component such as proper protonation and proper partial charges assignments.
Good Luck.
Bog
Zhaoxi Sun and Boguslaw Stec
Thank you very much. I will keep in mind your advices and suggestions.
Best wishes!
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Atomic, Molecular and Optical Physics
- Atoms