You are here

Rna Denovo

6 posts / 0 new
Last post
Rna Denovo

Dear all,

Im using RNA Denovo rosetta protocol to build a 3D model of a small RNA segment (~10 nucleotides). For a specific purpose, i would like to model a ensemble of this ssRNA segment avoiding base pairing. I tried to modify some parameters but the ssRNA built is always packaged, with some base pairs. Is it possible to model such ssRNA using Rna Denovo protocol? 


Thanks in advance for any help!

Post Situation: 
Wed, 2020-05-06 08:40

Great question. It's really hard to avoid base pairing artificially, because it's so favorable! We have no direct facility to prevent base pairing, but one artificial method could be to model in chemically modified nucleotides like 3-methyl uridine (which you would specify in your fasta file as X[UR3]) or 1-methyl guanosine (X[1MG]) to prevent ideal pairs from forming. But even with ideal pairing prevented, you are still likely to get fairly "packaged" architectures, simply because minimizing the radius of gyration is energetically favorable as well. Can you share a bit more about the theoretical motivation here? It's possible that we can get at your goal another way, or that this is simply the wrong modeling tool for your needs.

Wed, 2020-05-06 15:47

Hi! Thanks for your response! In my structural biology problem i have a target protein which has a long cleft that receives ssRNA (6-10 nucleotides). Despite the many degree of freedom inherent in that RNA, i was trying to produce a less packaged ensemble of ssRNA structures to docking. 


Thu, 2020-05-07 05:16

Got it. It sounds like you need to encode that specific constraint a bit! Are you familiar with DRRAFTER? This code allows you to take a structure of your protein, electron density, and model the ssRNA in. (Even if you don't have electron density, we have a capacity to fold-and-dock RNA into protein binding pockets.) See and

Thu, 2020-05-07 10:24

Thanks! I'll take a look at DRRAFTER. I have already tried the RNP-denovo fold-and-dock method. By the way, do you know if it is possible to provide protein binding site information (for restraints for example) in that fold-and-dock method? 

Fri, 2020-05-08 13:09

Hey, sorry I wasn't around. You can definitely provide protein binding site information using restraints. The code can respect your command line constraints file, and there's plenty of documentation surrounding how to provide (say) AtomPair constraints using your own function definition that has the profile you want. (Say, no penalty as long as the two atoms are within 10 angstroms of each other -- you can be pretty generous if you need.)

Wed, 2020-05-20 08:52