Question: What Tool To Use For Protein-Rna Docking?
6
gravatar for Cosi
9.3 years ago by
Cosi60
Gliwice, Poland
Cosi60 wrote:

Hello, A field of my research is RNA interference. I'd like to examine its efficiency after stress-induced damage in silico. The experiment would be performed in the following steps: 1) Obtaining structure files of a protein domain (PAZ, PIWI, Mid) and small RNA 2) Docking computations 3) Modification of protein/RNA structure and repeating step 2 4) Comparison of the results. Actually, I'm new to molecular modeling and have no idea what tools I should use. I need something that will allow me to change a single amino acid/nucleotide or oxidize nucleotides. Would AutoDock+ADT be enough for this job?

PS. I'm a newbie, so please be tolerant :)

protein rna docking • 8.7k views
ADD COMMENTlink modified 3.8 years ago by liushiyong0 • written 9.3 years ago by Cosi60
7
gravatar for Jan Kosinski
9.3 years ago by
Jan Kosinski1.6k
Jan Kosinski1.6k wrote:

1) Protein-RNA docking - the field is in infancy, docking more rigid RNAs that do not change much after binding to the protein may succeed, but flexible RNAs, even small, would most likely fail. If you google a bit, there is a considerable information about protein-RNA docking already. For example, you easily can find that protein-RNA docking has been included in CAPRI (molecular docking contest) and it failed when unbound RNA structure was used.

2) RNA modification - if you want to oxidize nucleotides, you can try a tool from my previous lab: http://genesilico.pl/moderna/ "ModeRNA is capable of handling 115 different nucleotide modifications"

3) Protein in silico mutagenesis - I would use Rosetta http://www.rosettacommons.org/software/ together with http://www.pyrosetta.org/

4) There is little hope that your experiment will give any conclusive results - as I understand you want to dock RNA to protein and check how little changes to protein or RNA affect the result. I think that uncertainties in the docking results themselves are too big to analyze the effect of fine changes like single point mutants etc. The uncertainties are even bigger as it's still not straightforward to predict the effects of mutations on the structure.

I was docking dsDNA a lot, both by comparative modeling and de novo docking, and even in comparative modeling (much more accurate than de novo modeling) there were significant changes in several prot-DNA interactions due to even slight changes in DNA angles. No hope that in silico I could assess quantitatively the effects of single point mutants.

In one of the projects (http://www.ncbi.nlm.nih.gov/pubmed/17407166), I was also trying to explain the known effects of small DNA modifications based on protein-DNA model, and only some explanations were correct as revealed later by crystal structure (http://www.ncbi.nlm.nih.gov/pubmed/17344322)

ADD COMMENTlink written 9.3 years ago by Jan Kosinski1.6k

Thanks for your explanations. That's a bit disappointing, but I'll try to find the way to do it in silico (maybe using machine learning methods based on existing data?) nonetheless, since I have no access to X-ray or NMR in Gliwice.

ADD REPLYlink written 9.3 years ago by Cosi60

Start with analyzing well known structures of your protein domains in complex with RNA. Perhaps you can find some rules, regular patterns of binding modes, etc. Perhaps the complexes are so well characterized that you don't need docking at all.

And define what answers you expect to get in the best case, and what anybody could do with that data. It's easy to do a project in bioinformatics, more difficult is to provide any useful answers.

ADD REPLYlink written 9.3 years ago by Jan Kosinski1.6k

one of the main problems here is that no method treats efficiently macromolecular flexility for the whole molecule (i.e. auto dock)

ADD REPLYlink written 8.7 years ago by Flow1.5k
5
gravatar for João Rodrigues
8.7 years ago by
João Rodrigues2.5k
Stanford University, U
João Rodrigues2.5k wrote:

Probably a bit too late, but HADDOCK webserver (http://haddocking.org) does accept RNA as a molecule type. It is also suitable if you have experimental information that could help in the docking process such as mutagenesis, or just interface predictions from any bioinformatics server. Maybe worth giving a try if you are still up to it.

ADD COMMENTlink written 8.7 years ago by João Rodrigues2.5k
2
gravatar for Chris Evelo
9.3 years ago by
Chris Evelo10k
Maastricht, The Netherlands
Chris Evelo10k wrote:

I have never done this. But I would have a look at [?]WhatIf[?]. It is designed to ask questions like "what would happen if I replace one amino acid". But I am not sure how well it handles RNA.

ADD COMMENTlink written 9.3 years ago by Chris Evelo10k

Thanks, Chris. It looks interesting. It's a pity that there's no demo version to download...

ADD REPLYlink written 9.3 years ago by Cosi60

You are right. I checked. Sorry, I wasn't aware that there is a fee for academic users. I think it might be worth to mail the author and inquire about that. Maybe you can in fact use a demo version and you can also ask about what it can do with RNA. Although the latter might also be in the manual (click WriteUp)

ADD REPLYlink written 9.3 years ago by Chris Evelo10k
0
gravatar for liushiyong
7.1 years ago by
liushiyong0
liushiyong0 wrote:

You may try RPDock.

http://www.nature.com/srep/2013/130528/srep01887/pdf/srep01887.pdf

http://biophy.hust.edu.cn/3dRPC.html

ADD COMMENTlink written 7.1 years ago by liushiyong0
0
gravatar for liushiyong
3.8 years ago by
liushiyong0
liushiyong0 wrote:

You may try PRIME

http://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1005120

http://rnabinding.com/PRIME.html

ADD COMMENTlink written 3.8 years ago by liushiyong0
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1762 users visited in the last hour