Hi,
During a computational work, I got some interesting result. I found a strong interaction between two proteins from Molecular dynamics simulation. A 100ns simulation clearly shows the stability of complex and stable H-bonding pattern. However, the two published literature denied their direct interaction. I do feel this is a interesting finding, but I ain’t getting it published despite of various attempts. Please suggest me some journals.
Dear Jeetandra,
It is indeed a very interesting question. I am not sure whether I would be any help but I have been in a similar situation so I would post an answer.
If one gets a negative result and it does not have a precedence then the article finds it tough in the publishing house. I tried to compare my observation to experimental results. Although, the scales of my simulation and the experiment (from literature) were different, I attempted some scaling analysis and tried to establish a relation. Some of the effects were found to be extremely scale dependent and after some iterations in various journals it finally got published. 3 years down the line, the concept of shear dependent capillary slip introduced in the paper is well accepted and there have been more publications from different groups.
You do some great research and probably this response gives you no new information. However, these are the few things you can try.
1) Compare your simulation setup with the reported literature. They might have neglected something or there is some additional parameter. If you find one that parameter induces the protein interactions.
2) Look for experiments which report similar interactions and that could help to establish your point.
3) This is weird. Try to get examples from direct biological/living systems,
Regards,
Chirodeep.
What was the starting structure of your simulation? What you have to keep in mind is that MD simulations are prone to being stuck in local minima, and 100 ns is certainly not enough time to get out of that minimum, in conjunction with complications associated with potential force field bias. If the initial structure is docked or from a crystal, these may have no relevance to anything done in solution. Experimental time scales are much longer than what we can generally explore with simulations, which is a limiting factor.
#Many thanks Dr Bakli.
#Jastin,
The primary structures considered for the study are crystal structures. The electrostatic surfaces clearly indicates the possible interaction (what we got). During MD studies of the complexes becoming stable after 50ns of MD and a stable H-bonding pattern. The inter-atomic distance calculation also shows stable individual H-bonds.
Crystal complexes and solution behavior are almost always very different, since what is possible to co-crystallize is not necessarily what happens in solution. The fact that the electrostatic surface yields a strong interaction does not tell you anything really conclusive in this regard, because in solution, to form a complex, the interface residues have to be desolvated in order to come together as they are in the crystal. There is going to be a large energy penalty for doing this, which is often overcome by somewhat harsh and dehydrating conditions in crystallization buffers.
If you were to run a simulation starting from separate monomers, and those came together, and you could make an actual free energy argument (e.g. PMF) as to why this is happening, that would be compelling. But as it stands, you're just sitting in an energy minimum that likely has little or no relevance to what happens in solution. The inherently over-polarized amino acid side chains inherent to additive force fields are going to bias the result towards this outcome.
If your hypothesis is "the two proteins DO interact with each other" then your null hypothesis must be "the two proteins DO NOT interact with each other". If your calculations provide strong enough "evidence" for invalidating/challenging/doubting the null hypothesis then you get a POSITIVE result.
- Badges
- Science topic