Mutate_Bases: Option to mutate all residues of the same type to another

[slaw]

Xiang-Jun,

I am playing with mutating 1RNA.pdb (contains only A/U) into a structure that contains only C/G (G for A and C for U).  This is not vital but it would be nice if there was an option to convert ALL of the A's to G's (or whatever is needed) and all of the U's to C's.  The current functionality is perfectly fine for a few point mutations but this added capability would be helpful and make the process less tedious having to create a file to list out all of the individual mutations.  I can also see a logical negation to be useful as well.  For example, mutate all of the A's to G's except for residue 5 and residue 7.

Again, the program works great so this would only be a wish-list request.

Thank you for your time, effort, and continued support of 3DNA!

[xiangjun]

Thanks for your request for adding new options to mutate_bases "to mutate all residues of the same type to another". I see your point, but I feel such functionality is better served by purpose-specific scripts instead of by mutate_bases itself. As a general design principle, I aim to make the core part of 3DNA focused, robust and efficient, and use Ruby scripts to build up features of general interest. Overall, 3DNA is intended as a toolset to be used as users see fit.

With regard to mutate_bases, I am considering to add an option to list all bases (nucleotides) in a PDB structure in a format that can be easily modified (either manually or pragmatically) and then fed back to the program. This option allows for easy adaptation of mutate_bases to various common situations (including yours), without adding too many bells and whistles that would blur the program's primary function.

An excerpt example for the first 12 bases of 6tna is as below:

Code: [Select]

chain=A snum=1 name=G
chain=A snum=2 name=C
chain=A snum=3 name=G
chain=A snum=4 name=G
chain=A snum=5 name=A
chain=A snum=6 name=U
chain=A snum=7 name=U
chain=A snum=8 name=U
chain=A snum=9 name=A
chain=A snum=10 name=2MG
chain=A snum=11 name=C
chain=A snum=12 name=U
Does this sound useful?

Xiang-Jun

[manish]

Hi Xiangjun,

My question is related to mutate_bases option in the x3DNA webpage. When I select any pdb file for mutation, it gives me the option of mutating bases like A, T, G, C or U at any position in the selected pdb file. Is it possible to include rNMP apart form dNMP in these mutate_bases option even if my control sequence is composed of dNMP units only?  So basically, i want to keep the base part unchanged but the sugar can be changed from deoxyribose to ribose (Eg- dA to rA or dG to rG as such) with the predefined fiber models.

Thanks a lot for your time

Manish

[xiangjun]

Hi Manish,

Thanks for trying out the new web 3DNA 2.0. As the name mutate_bases hints, the program mutates bases only while keeping the backbone untouched. So it is (currently) not possible to change the sugar from deoxyribose to ribose.

In 3DNA v3.0 (where DSSR and SNAP are two representatives), I will consider adding more modeling features based on users' requests, on a case-by-case basis. In this specific case, aren't there other software tools available for sugar mutations? Cannot you do it manually? How do you consider the sugar conformations (C3'-endo vs C2'-endo) after adding O2'? Do you have a concrete example of such mutations?

Best regards,

Xiang-Jun

[manish]

Hi Xiang-Jun,

First of all, thanks a lot for your earnest response. With regards to your first query, I am not sure if there are any available software tools for sugar mutations. However, please let me know if you find/are aware of any such tools for sugar mutations. It would be of great help to me.

' Cannot you do it manually? '- I can do it manually but i will have to worry about the dihedral angles and bond length in the case of the mutated sugars and it might be inconsistent with the conformation that i have taken from the fiber model from web x3dna 2.0. So, O2'-C2'-C3'-C4' dihedral is the main cause of my concern.

' How do you consider the sugar conformations (C3'-endo vs C2'-endo) after adding O2' ?'- Ideally, I would like to have the sugar conformations unchanged for the modified bases after addition of O2', if there are no steric clashes. Then, I want to see what changes come up over time to these forcibly retained bases, through MD simulations.

' Do you have a concrete example of such mutations? '- Yes, there are many examples of such mutations in literature. Please go through either of these links to have a rough idea of what I am trying to study:
1- https://onlinelibrary.wiley.com/doi/full/10.1002/cbic.201600385
2- https://pubs.acs.org/doi/10.1021/bi201710q

I would be glad to listen to any of the ideas you have to make these modified sugars. Thanks again for your time and cooperation.

 

[xiangjun]

Hi Manish,

Thanks for your detailed responses to my follow-up questions. Presumably, such a back-and-forth conversational style should be the norm on an online forum. In practice, as one can see by browsing the 3DNA Forum, that is normally not the case. I sense users may just want a quick and straight answer to their questions, which are frequently not well defined. I then ask for clarifications, but oftentimes receive no responses.

I am not sure if there are any available software tools for sugar mutations. However, please let me know if you find/are aware of any such tools for sugar mutations. It would be of great help to me.

I'm aware of the paper "Using internal and collective variables in Monte Carlo simulations of nucleic acid structures: Chain breakage/closure algorithm and associated Jacobians" by Heinz Sklenar et al. which may be relevant to the topic here.

' Do you have a concrete example of such mutations? '- Yes, there are many examples of such mutations in literature. Please go through either of these links to have a rough idea of what I am trying to study:
1- https://onlinelibrary.wiley.com/doi/full/10.1002/cbic.201600385
2- https://pubs.acs.org/doi/10.1021/bi201710q

I will follow these two links and see what I can get. Since your request in on sugar mutations which is not covered by the 'mutate_bases' program in 3DNA v2.x, please start a new thread for future communications.

Best regards,

Xiang-Jun

[xiangjun]

As a follow-up of the initial request "Mutate_Bases: Option to mutate all residues of the same type to another". In 3DNA v3, I've added a new program named x3dna-mutate which can do this and much more, by incorporating the DSSR annotations. Some examples:

Code: [Select]

x3dna-mutate --mutation='to=A' -i=1ehz.pdb -o=1ehz-allA.pdb
x3dna-mutate --mutation='name=A to=G' -i=1ehz.pdb -o=1ehz-A2G.pdb
x3dna-mutate -m='to=A cond=hairpin' -i=1ehz.pdb -o=1ehz-hairpinsA.pdb
x3dna-mutate --m='name=A negate to=G' -i=1ehz.pdb -o=1ehz-nonA2G.pdb
x3dna-mutate --m='chain=A num=12 to=DT' -i=355d.pdb -o=355d_G12T.pdb
x3dna-mutate --m='num=12 to=DT; num=13 to=DA' -i=355d.pdb -o=355d_TA.pdb
Xiang-Jun

[xiangjun]

As a follow-up of this thread, DSSR 2.0 (to be released soon) contains a new module for in silico base mutations with great flexibility and convenance. It can mutate all of the A's to G's except for residue 5 and residue 7, for example.

The 'mutate_bases' program is obsoleted. No program named 'x3dna-mutate' will be distributed.

Xiang-Jun