Missbonding in unpaired region

[yuxiao zhang]

Hi Sam,

I've been modeling a complex composed of 3 ssDNA, here is the code I use:

#sequence
DNA A 1 AAAAGCTGTGACTCGT
DNA B 1 AAAACCCCCAGGTTCTCT
DNA C 1 ACGAGTCACAGCAGAGAACCTGGGGGAGTATTGCGGAGGAAGGT

# stages
firstStage 1
lastStage 1

#general simulation parameters
# This is a multiplying factor for the baseInteraction forces:
baseInteractionScaleFactor 200
# Simulation time = reportingInterval * numReportingIntervals :
reportingInterval 4.0
numReportingIntervals 10
temperature 10

baseInteraction A 16 WatsonCrick C 1 WatsonCrick Cis
baseInteraction A 15 WatsonCrick C 2 WatsonCrick Cis
baseInteraction A 14 WatsonCrick C 3 WatsonCrick Cis
baseInteraction A 13 WatsonCrick C 4 WatsonCrick Cis
baseInteraction A 12 WatsonCrick C 5 WatsonCrick Cis
baseInteraction A 11 WatsonCrick C 6 WatsonCrick Cis
baseInteraction A 10 WatsonCrick C 7 WatsonCrick Cis
baseInteraction A 9 WatsonCrick C 8 WatsonCrick Cis
baseInteraction A 8 WatsonCrick C 9 WatsonCrick Cis
baseInteraction A 7 WatsonCrick C 10 WatsonCrick Cis
baseInteraction A 6 WatsonCrick C 11 WatsonCrick Cis
baseInteraction A 5 WatsonCrick C 12 WatsonCrick Cis
baseInteraction B 18 WatsonCrick C 13 WatsonCrick Cis
baseInteraction B 17 WatsonCrick C 14 WatsonCrick Cis
baseInteraction B 16 WatsonCrick C 15 WatsonCrick Cis
baseInteraction B 15 WatsonCrick C 16 WatsonCrick Cis
baseInteraction B 14 WatsonCrick C 17 WatsonCrick Cis
baseInteraction B 13 WatsonCrick C 18 WatsonCrick Cis
baseInteraction B 12 WatsonCrick C 19 WatsonCrick Cis
baseInteraction B 11 WatsonCrick C 20 WatsonCrick Cis
baseInteraction B 10 WatsonCrick C 21 WatsonCrick Cis
baseInteraction B 9 WatsonCrick C 22 WatsonCrick Cis
baseInteraction B 8 WatsonCrick C 23 WatsonCrick Cis
baseInteraction B 7 WatsonCrick C 24 WatsonCrick Cis
baseInteraction B 6 WatsonCrick C 25 WatsonCrick Cis
baseInteraction B 5 WatsonCrick C 26 WatsonCrick Cis
baseInteraction C 31 WatsonCrick C 41 WatsonCrick Cis
baseInteraction C 32 WatsonCrick C 40 WatsonCrick Cis
baseInteraction C 34 WatsonCrick C 38 WatsonCrick Cis

There is a problem that bonds of unpairing parts are unrealistic while the duplex region is OK.

Missbonding.png (198.69 KiB) Viewed 2264 times

I can't figure out why this happens because I'm a beginner. It would be great to know what command should I take to deal with this?

Thanks!
Yuxiao

[Samuel Flores]

In the unpaired region there are not restraints at all, so the residues can do as they wish.

I would start by adding collision detecting spheres, use the AllHeavyAtomSterics scheme.

For the helical parts, you will get better quality if you use NtCs, if you don't know you can assume you need AA00. I believe you will find examples of all of this in the tutorial guide. Are you using our latest version?

Sam

[yuxiao zhang]

Thanks for your kind instruction!

I first tried to add MD command for the unpaired region:

setDefaultMDParameters
includeResidues A 1 4
includeResidues B 1 4
includeResidues C 27 44

The problem of miss bonds in the unpaired region was solved.

Then I rigidified the unpaired region to run MD for the paired region, but the energy was extremely high. After several attempts, I found that the problem came from base 12 and 13 on chain C.

NUPACK.png (79.45 KiB) Viewed 2242 times

I added NtC C 12 13 OP03 .5 to the command which generate a structure seemed to be OK.

I'm using MMB 3.0 now.

Best
Yuxiao