Messages

1401

MD simulations / Re: Ruby scripts for the analysis of MD simulation trajector

« on: June 08, 2011, 09:36:23 pm »

Hi Henry,

Thanks for using 3DNA. I fully understand how frustrated it could be to get a new software system up and running, and you are welcome to ask any 3DNA-related questions in the forum.

The Ruby scripts may be useful if you are using AMBER for MD simulations. For building general sequence single strand RNA structures, the recent post "building a ssDNA" may help. Run also "fiber -m" to see all available fiber RNA models.

Regarding the specific problem you have with running the Ruby script '[mono:1bnxcgla][red:1bnxcgla]x3dna_md.rb[/red:1bnxcgla][/mono:1bnxcgla]':
[pre:1bnxcgla]3DNA settings error: can't execute C:/MinGW/msys/1.0/home/bshcf/X3DNA/bin/analyze[/pre:1bnxcgla]
It is likely due to line #425:
[pre:1bnxcgla]fatal("3DNA setting error: can't execute #{x}") unless File.executable?("#{x}")[/pre:1bnxcgla]
change it to (i.e., by adding [mono:1bnxcgla][red:1bnxcgla].exe[/red:1bnxcgla][/mono:1bnxcgla] at the end since you are using MinGW)
[pre:1bnxcgla]fatal("3DNA setting error: can't execute #{x}") unless File.executable?("#{x}[red:1bnxcgla].exe[/red:1bnxcgla]")[/pre:1bnxcgla]
Have a try, and report back how it goes.

Generally speaking, the best way to get started with 3DNA is by reading the 2008 3DNA Nature Protocols paper, and working out the recipes.

HTH,

Xiang-Jun
[hr:1bnxcgla][/hr:1bnxcgla]
Since this thread is becoming too long, please start a new one with a specific Subject line.

1402

General discussions (Q&As) / Re: change one base pair in a double-strand DNA structure fi

« on: June 04, 2011, 11:16:50 pm »

Hi all,

I have finally made the programs ready for distribution. There are actually two variants as detailed below for mutating bases in a nucleic acid structure while keeping the sugar-phosphate backbone untouched. Both the mutate_bp Perl script for base-pair mutations and the more generic/efficient/robust ANSI C program mutate_bases possess the neat and indeed unique feature, i.e., the mutated structures share the same base-pair/step parameters as the original structure.

• mutate_bp is a Perl script I first wrote a couple of years ago. Its usage is as follows:
  

  mutate_bp -m=bpNUM_BP input.pdb output.pdb
  
  The mutation string must be in the format of 'bpNUM_BP' where NUM is
      an integer for the base-pair to be mutated (check 'find_pair'),
      and BP is the desired base-pair. BP must be two letters from the
      set ACGTU; e.g. AT/TA/AU/UA/GC/CG for Watson-Crick pairs. Case
      does not matter (i.e., 'at' is the same as 'AT').
  
  e.g. mutate_bp -m=bp2_at 355d.pdb 355d_bp2at.pdb
         # mutate the second base-pair, G-C, to A-T in PDB entry 355d

  
  This script presumably should fulfill Chien-Ho's need, "change one base pair in a double-strand DNA structure file", as well as the two previous similar requests on this topic. Technically, the script makes (system) calls of various 3DNA (v2.0) programs (e.g., find_pair, analyze, rebuild, frame_mol etc.), and basically follows the procedures I outlined in response to "mutating DNA in DNA protein complex" (Oct. 2007).
  
  The mutate_bp Perl script is directly available for download. Simple save it to your $X3DNA/bin directory, and make sure it is executable (run chmod a+x mutate_bp if necessary). Interested users are encouraged to explore the source code and make changes as they see fit.

• mutate_bases is a standalone ANSI C program, on a par with other major 3DNA programs (e.g., find_pair, analyze, rebuild, fiber etc). As seen from the help message below, it can be used for any nucleic-acid-containing structures (DNA, RNA, or their complexes, in PDB format):
  

  Usage: mutate_bases mutinfo pdbfile outfile
     'mutinfo' can contain upto 5 fields for each mutation:
         [name=residue_name] [icode=insertion_code]
         chain=chain_id seqnum=residue_number
         mutation=residue_name
     o The five fields per mutation can be in any order or CaSe
     o Each field can be abbreviated to its first character
     o Multiple mutations are separated by ';'
     o Fields in [] (i.e., name and icode) are optional
     o Mutation info should be QUOTED to be taken as one entry
  
  Examples:
      mutate_bases 'c=a s=2 m=DA' 355d.pdb 355d_G2A.pdb
     # mutate G2 in chain A of B-DNA 355d to A
      mutate_bases 'c=a s=2 m=DA; c=B s=23 m=DT' 355d.pdb 355d_GC2AT.pdb
     # mutate base-pair G-C (C23 in chain B) to A-T
      mutate_bases 'c=A s=74 m=U' 1evv.pdb 1evv_C74U.pdb
     # mutate C74 in chain A of tRNA 1evv to U

  
  mutate_bases is designed to solve the base mutation problem in a practical sense: robust and efficient, getting its job done and then out of the way. The program can have many possible applications: in addition to perform base-pair mutations in DNA-protein complexes as mutate_bp does, mutate_bases should prove handy in providing initial structures for QM/MM/MD energy calculations, and in RNA modeling studies.
  
  While mutate_bases is now fully functional, it is not intended for general release yet. Interested users should contact me directly for access to the program (Linux, Mac OSX Intel, and MinGW-MSYS binary versions only; 3DNA v2.0 required).

As always, I welcome users feedback; any questions and comments should be posted in this forum.

Xiang-Jun

1403

General discussions (Q&As) / Re: change one base pair in a double-strand DNA structure fi

« on: May 27, 2011, 10:55:52 pm »

Although the two topics in this forum could be well solutions of my problem, I still want to try the component "mutate_bases" for more experiments. Could you share the component with me please? The OS of my working server is Linux and my 3DNA is version 1.5.

Please upgrade your 3DNA to v2.0, by following "How to download 3DNA v2.0?". The program "mutate_bases" needs 3DNA v2.0 to run. For outside use, I am planning to clean up the code and test it further. Stay tuned ... hopefully I will be able to get the job done by early June.

Xiang-Jun

1404

General discussions (Q&As) / Re: change one base pair in a double-strand DNA structure fi

« on: May 25, 2011, 09:43:25 pm »

Hi Chien-Ho,

Thanks for using 3DNA, and for posting at this forum.

The question on mutating a DNA base (or base-pair) while keeping the backbone untouched has showed up a couple of time in this forum:

• Problems with mutations on DNA (Feb. 2011)
• mutating DNA in DNA protein complex (Oct. 2007)

Follow the info there, and hopefully, your problem can be solved.

On the other hand, I have added a new component to 3DNA, named "mutate_bases" that does exactly this job. Moreover, it can be applied to RNA structures as well. The program is written in ASNI C, and is commandline-driven, as other major 3DNA components. It has been used internally for a while, and works as designed. If you or others would like to have a try, please post back here. Let me know what OS you use. I will consider to make "mutate_bases" partially available for 'testing' purpose before I include it in a later major 3DNA release (no date specified).

HTH,

Xiang-Jun

1405

General discussions (Q&As) / Re: Sequence dependent bp step parameters

« on: May 19, 2011, 10:38:41 pm »

Hi Søren,

Thanks for your 3 posts, and I wish I could have been able to get back to you earlier. However, w3DNA is out of my reach: I am not that familiar with the internal details of w3DNA, no access to the server, no time to devote to it (unfortunately).

3DNA itself does not have "default" bp step parameter values for rebuilding. The "B-DNA (generic)" option you mentioned may refer to the fiber models. To verify, run "[mono:3nnrwi6s]fiber[/mono:3nnrwi6s]" to generate a B-DNA (default) model, and then "[mono:3nnrwi6s]analyze[/mono:3nnrwi6s]" it to see the parameters involved. Moreover, 3DNA does not have a compilation of various bending models (as SCHNArP does), so any comparison with other parameter sets may not be meaningful.

Xiang-Jun

1406

General discussions (Q&As) / Re: Base pair step parameters with a missing base pair

« on: May 03, 2011, 11:43:26 pm »

Hi Mithun,

The following info may help:

• 3DNA FAQ "How to fix missing (superfluous) base pairs identified by find_pair?" You may want to upload (as attachment) the 4-mer duplex fragment (residues 96-98 for one strand) and provide the base pairs found with "find_pair". Then we can discuss specifically why some pair is missed or mis-assigned by "find_pair".
• Search the forum using the term "find_pair".
• Since you are using 3DNA for MD simulations, please see the post "Ruby scripts for the analysis of MD simulation trajectories".

HTH,

Xiang-Jun

1407

General discussions (Q&As) / Re: Phi-angle problem?

« on: May 02, 2011, 09:38:24 pm »

Hi Søren,

Thanks for your interest in SCHNAaP/SCHNArP and 3DNA. I am impressed by your diligence in digging into the bottom of seemingly trivial details; in my experience, they are what count the most.

Regarding the definition of the phi-angle, reading El Hassan and Calladine's original JMB paper, and my SCHNAaP/SCHNArP JMB publications would certainly help. However, they are no substitutes of going through a worked example -- that's exactly what I put into Section #5 of the 3DNA user's manual (http://3dna.rutgers.edu:8080/3DNA_v1.5/x3dna.pdf). In 3DNA v2.0 distribution, I extracted that section into a separate file named "tech_details.pdf". For any serious 3DNA users, I strongly suggest them to work through it carefully in order to understand each step thoroughly. If you want to see my SCHNAaP/SCHNArP implementation in ANSI C, please let me know.

I am glad that you "found the problem". Regarding "a typo in the SCHNArP paper equation 21" that confused you, I agree that there is indeed a typo -- thank for bring this issue to me attention:
[pre:2wbo9dyp]T^g_{i+1} = [red:2wbo9dyp]T^T_i[/red:2wbo9dyp] * T^i_{i+1}   -- (21) from the SCHNArP paper
should be
T^g_{i+1} = [red:2wbo9dyp]T^g_i[/red:2wbo9dyp] * T^i_{i+1}[/pre:2wbo9dyp]
 I.e., the [red:2wbo9dyp]red[/red:2wbo9dyp] term should have a superscript [red:2wbo9dyp]g[/red:2wbo9dyp] instead of [red:2wbo9dyp]T[/red:2wbo9dyp]. Note that two terms following the equal sign is the correct order.

There are many subtle details that must be properly accounted for in order to get the correct result. Note that the equations may be derived alternatively in a mathematically equivalent form; this may explain why you have to swap the two right hand terms in equation 21 to solve your puzzle.

HTH,

Xiang-Jun

1408

General discussions (Q&As) / Re: Rebuild DNA from C trace

« on: April 20, 2011, 08:51:14 pm »

Dear Damien,

Thanks so much for sharing your story, and I am glad to know that your problem has been solved with  YASARA. I heard of YASARA before, but did not know that it can be used to build a whole DNA structure with only C trace.

Your question inspired me to think a bit more about the DNA modeling issue starting from a partial C or P trace. I sense the solution must be approximate, and not unique. If I have the luxury to devote more effort to 3DNA than my spare time could afford, I may come up with a geometry-based command-line-driven tool. Of course, if YASARA has solved this problem well enough, why bother to recreate the wheel.

Good luck with your project.

Xiang-Jun

1409

General discussions (Q&As) / Re: Rebuild DNA from C trace

« on: April 15, 2011, 07:08:48 pm »

Hi Damien,

Thanks for posting at the 3DNA forum. Unfortunately, 3DNA (in its current version) cannot rebuild a whole DNA structure from only C4 trace. Off the top of my head, I am not sure of another tool that can perform the task either :oops:. You may have a look of NAB from David Case's lab.

Hope you would post back how it goes, possibly with more background information. Whenever possible, I am always interested in improving 3DNA in ways that make sense to me and helpful to the community.

Xiang-Jun

1410

General discussions (Q&As) / Re: building a ssDNA

« on: April 10, 2011, 12:31:42 pm »

But this dna doesn't have asymmetric ends, while I need to cap it with 5' and 3'. Can you guide me how do make the ends 5', and 3'?

What do you mean "DNA capped with 5' and 3' ends"? Please be specific, illustrated with examples.

Do you see such caps with DNA structures downloaded from PDB/NDB?

Again, I sense this concept may be more related to the requirements of your MD simulations than to 3DNA.

Xiang-Jun

1411

MD simulations / Re: Ruby scripts / where is output file?

« on: April 08, 2011, 07:56:14 pm »

Did you read my reply dated Apr 02? Do you know that the Ruby scripts are built on top of 3DNA -- i.e. you should already have a valid 3DNA (preferably v2.0) installed? When you said that "unfortunately in my output file (my.out) the value of parameter relating to every base pair is the same during trajectory", did you bother to check which base-pair is being repeated, and wondering why that bp specifically, not another one? Without being able to reproducing your problem, I really cannot offer you any concrete suggestion. Please ask a local expert (in Unix) for help. In any event, you have full access to the Ruby scripts to dig to the bottom of the problem.

Also as suggested in my previous reply (Apr 02), please update to the current version [red:3pnrsf1p]v0.5[/red:3pnrsf1p], follow carefully the instructions in the [red:3pnrsf1p]README[/red:3pnrsf1p] file, and report back explicitly what happens. The v0.5 scripts are designed to check more vigorously their prerequisites and complain loudly if the requirements are not met.

HTH,

Xiang-Jun

1412

General discussions (Q&As) / Re: building a ssDNA

« on: April 08, 2011, 07:39:46 pm »

Hi Majid,

Glad to see that with 3DNA v2.0, you can now create single stranded DNA. Regarding the topology file and the error message you referred to, they clearly have nothing to do with 3DNA; I sense they may be related to the Gromacs molecular dynamics package you referred to in your first message of the thread.

HTH,

Xiang-Jun

1413

General discussions (Q&As) / Re: How to determine DNA topology form parameters

« on: April 06, 2011, 09:57:18 pm »

Hi, I am analyzing a protein-DNA complex in which DNA is severely curved. The protein could constrain negative DNA supercoils in vitro and I am trying to find some clues to the structural basis of DNA supercoiling from the parameters of DNA calculated by 3dna (listed as following).

Is it possible to infer whether the supercoils come from DNA unwinding or writhing (e.g., the DNA fragment in nucleosome) from these parameters? If so, how to make it?

Thanks for using 3DNA. Your question "How to determine DNA topology form parameters" seems to be beyond the scope of 3DNA (at least in its current versions). In my understanding, DNA topology is at a "higher" level than can be directly accounted for by the base-pair and dinucleotide structural parameters calculated by 3DNA.

And the DNA has an average h-twist value of 35.29 and thus a h=(360/35.29)=10.20 bp/turn. Does it means that the DNA is over-twisted, since h is less than 10.5 bp/turn?

I would be very cautious in trying to draw any firm conclusion from such a subtle difference in helical twist. For one thing, the widely cited empirical value of 10.5 bp/turn is based on (presumably) B-form DNA in solution. The DNA in your protein-DNA complex is only 8-bp long and is severely curved.

Since you [red:3v097b1n]highlighted in red of twist and h-twist[/red:3v097b1n] in the enclosed 3DNA output file, I think the following two threads in the forum may help clarify confusions 3DNA users often have with the two and other related parameters:

• h-twist vs. twist (Oct 2006)
• local helical parameters interpretation question (Sep 2008)

In connection with DNA supercoiling, there is yet another twist in the twist parameter: please see "Two perspectives on the twist of DNA." by Britton, Olson & Tobias (J Chem Phys. 2009 Dec 28;131(24):245101).

HTH,

Xiang-Jun

[hr:3v097b1n][/hr:3v097b1n]PS: To the extent 3DNA may be useful to you, please consider to update your copy to 3DNA v2.0 -- it is simply a better version than v1.5!

1414

MD simulations / Re: Ruby scripts for the analysis of MD simulation trajector

« on: April 04, 2011, 09:11:09 pm »

Hi Alpay,

Glad to hear your verification that "the scripts are working nicely". I am well aware of the memory issue you raised. By design, the scripts currently process each and every snapshot of an MD simulation and collect all the parameters along the way. As to how many snapshots can be handled, the answer would obviously depend on computer memory and the size (i.e. no. of bps) of a snapshot. I would imagine the scripts should work well for an MD simulation with 50k snapshots,12 bps each, in a computer with say 4GB memory. As an estimation, the output parameters file for the sample 12 bps/20 snapshots takes 76,926 bytes. With 50k snapshots, the space taken would be  (76,926 / 20) * 50000 = 192,300,000 bytes, which is ~183 MB.

HTH,

Xiang-Jun

1415

MD simulations / Re: Ruby scripts / where is output file?

« on: April 02, 2011, 10:11:49 pm »

I was unable to reproduce your problem where in the output file "the value of parameter relating to every base pair is the same during trajectory".

As a followup of my reply (dated Wed Mar 09, 2011) to your post dated Tue Mar 08, 2011, did you realize that your base-pair file "bpfile.dat" does not match your PDB file "dna.pdb". Specifically, the first two base-pair lines in "bpfile.dat" read as below:

[pre:356ija09]1 [red:356ija09]28[/red:356ija09] 0 # 1 | A:...1_:[DC5]C-----G[DG3]:..26_:A 0.20 0.00 0.05 9.04 -1.30
2 [red:356ija09]27[/red:356ija09] 0 # 2 | A:...2_:[.DA]A-----T[.DT]:..25_:A 0.09 0.05 1.13 9.08 -1.30[/pre:356ija09]
However, your PDB data file "dna.pdb" has only 26 nucleotides. In other words, there are no nucleotides #27 and #28 in your PDB file at all.
[hr:356ija09][/hr:356ija09]
Over the past few weeks, I have used your example and other feedbacks I received to further refine the scripts. Now updated to v0.5, the scripts are much more robust;  they provide sensible error message if requirements are not met (e.g., with unmatched base-pair as in your case). All identified issues have been resolved.

Please have a try and report back how it goes.

HTH,

Xiang-Jun

1416

General discussions (Q&As) / Re: Conversions

« on: March 24, 2011, 09:54:20 pm »

Others and I may help you better if you are specific about the DNA duplexes created with "VMD and RasMol",  e.g., by attaching a sample file. Moreover, what format are those structures in initially? Is it PDB? If so, then why do you need to convert the files between .xyz and .pdb using OpenBabel?

By design, 3DNA only accepts compliant PDB format, as documented clearly in the PDB website. Please see the following two blog posts for more details:

• PDB format, how many variants are there?
• PDB ATOM coordinates record

HTH,

Xiang-Jun

1417

MD simulations / Re: Ruby scripts for the analysis of MD simulation trajector

« on: March 22, 2011, 09:33:45 pm »

Hi Alpay,

I have updated the Ruby scripts to v0.4 which fixed all the glitches that came to my mind.  Please have a try and report back how it goes.

Xiang-Jun

1418

MD simulations / Re: average values from MD simulations

« on: March 20, 2011, 10:50:35 am »

Hi Ara,

Welcome back.

So I was wondering if 3DNA can give information on which base exactly is fluctuating more causing the opening in the pair ?

No. 3DNA "can give information only on a base pair in this case".

if I can calculate the DNA bending during the simulation time

No.

or I will need to find an average structure of my simulation and calculate the bending of that structure.

It is entirely up to you. Please remember that 3DNA is just a tool set.

See my blog post "Calculation of DNA bending angle " for more information.

HTH,

Xiang-Jun

1419

General discussions (Q&As) / Re: building a ssDNA

« on: March 17, 2011, 10:36:28 pm »

Hi Majid,

Which version of 3DNA are you using? What '[mono:rzqs4j2j][red:rzqs4j2j]find_pair -h[/red:rzqs4j2j][/mono:rzqs4j2j]' gives you? Do you see the text "3DNA v2.0 [June 8, 2008]" at the bottom? If not, you are using a previous version of 3DNA (maybe v1.5), and it is high time to upgrade to v2.0.

The very simple Perl script '[mono:rzqs4j2j]pdb_frag[/mono:rzqs4j2j]' is included as part of 3DNA v2.0, in the [mono:rzqs4j2j]$X3DNA/bin[/mono:rzqs4j2j] directory. It is not documented because I do not think it worths the effort  :wink: -- simply look at the code will reveal everything. Given an opportunity, I will completely refine/extend the utility and then document it, with plenty examples   .

Please note the example I showed you is just an illustration of how single strand DNA model can be built using 3DNA, from the (limited) information you provided. There are a couple of other possibilities to explore as well. For those who really want to get the most of what 3DNA has to offer, the most effective way is to read carefully the 2008 3DNA Nature Protocols paper, and try to reproduce and understand each of the recipes/protocols -- I will be happy to address any issues users may encounter in repeating the results reported in this article.

HTH,

Xiang-Jun

1420

MD simulations / Re: Ruby scripts for the analysis of MD simulation trajector

« on: March 11, 2011, 07:46:15 pm »

From the error message, I guess you have some setting problems, e.g., mismatch between base-pair file and snapshots, or the snapshot contains unknown nucleotide that needs to be specified in file '[mono:2fyzp058]baselist.dat[/mono:2fyzp058]' etc. Thus, it is likely that no sensible 3DNA output parameters file was generated.

It is always a good idea to try a simple example with only one or two snapshots to verify that everything is set up correctly; afterwards, it is only a timing issue between 100 or one million snapshots.

Depending on user feedbacks, I am planning to refine the scripts to check common pitfalls and output sensible error messages accordingly. The next v0.4 should be much more robust for real world applications.

Xiang-Jun

1421

MD simulations / Re: Ruby scripts for the analysis of MD simulation trajector

« on: March 10, 2011, 09:29:42 pm »

Hi Alpay,

Thanks for the info. Clearly (but not surprisingly), there are many variants of ensemble-format from different visualization/simulation packages. It is impractical nor desirable for the Ruby scripts to account for each and every such possibility. By design, the scripts deal only the simplest/commonest case where each model/snapshot is delineated by the MODEL/ENDMDL pair, and are directly applicable to the analysis of NMR ensembles from the PDB/NDB. To make this point even clearer, I have added  the following sentence to the release post:

Importantly, for the [mono:2fyxsqwq]-e[/mono:2fyxsqwq] and [mono:2fyxsqwq]-m[/mono:2fyxsqwq] options, each model in the ensemble must be delimited by an MODEL/ENDMDL pair, as clearly documented in the Coordinate Section of the PDB format.

The commonly used ensemble formats are better and easily handled on a case-by-case basis, with the following three choices:

• Ask the original software package creator/maintainer to provide support for the MODEL/ENDMDL format, as it is well-documented in the PDB format.
• Write a purpose-specified script to perform the conversion as needed by the user.
• As time permits and with responsive and knowledgeable collaborators, I will consider to write and integrate conversion scripts into the distribution.

Note that starting with v0.1, the scripts support four options ([mono:2fyxsqwq]-e, -m, -p, -l[/mono:2fyxsqwq]) to allow great flexibility.

HTH,

Xiang-Jun

1422

MD simulations / Re: Ruby scripts / where is output file?

« on: March 09, 2011, 08:31:16 pm »

Since the scripts work (as designed) with the distributed example, yet not in your case,  please verify that your setting is correct. Specifically, make sure that your supplied base-pair file matches the model in your ensemble PDB file.

Please report back how it turns out.

Xiang-Jun

1423

General discussions (Q&As) / Re: building a ssDNA

« on: March 08, 2011, 12:05:59 am »

Hi Majid,

You may want to use '[mono:4xhr4aus]fiber[/mono:4xhr4aus]' to generate a duplex DNA of specific sequence, and then extract only chain A either manually, or with the simple Perl script '[mono:4xhr4aus]pdb_frag[/mono:4xhr4aus]'. Here is an example, [red:4xhr4aus]red[/red:4xhr4aus] text means user input:
[pre:4xhr4aus][red:4xhr4aus]fiber my_seq.pdb[/red:4xhr4aus]
Structure #4; Twist: 36.0 (degrees); Rise: 3.375 (Angstrom)

Input your base sequence with only A,C,G & T:
1. From a data file (complete sequence)
2. From keyboard (enter only the repeating sequence)
Your choice (1 or 2, Dft: 2): [red:4xhr4aus]ENTER[/red:4xhr4aus]    # to use the default option

Repeating unit (Dft: A): [red:4xhr4aus]ACGGGTTTTAAAAACCTTTA[/red:4xhr4aus]    # type your sequence
Repeating unit: ACGGGTTTTAAAAACCTTTA
Number of repeats (Dft: 10):  [red:4xhr4aus]1[/red:4xhr4aus]  # just repeat once of the sequence
# Now use RasMol or whatever your preferred viewer to display the duplex structure.

[red:4xhr4aus]pdb_frag A 1:20 my_seq.pdb my_ssDNA.pdb[/red:4xhr4aus]
# my_ssDNA.pdb may be what you want. Again, display it in RasMol.[/pre:4xhr4aus]
Of course, the above procedure could look cryptic and scary for GUI-driven point-and-click users. However, once you know what you are doing, it is easy to write a script to automate the process to ensure quick and reproducible computational 'experiments'.

Just play around to get more familiar with 3DNA. A good starting point could be to reproduce the reported results in the 2008 3DNA Nature Protocols paper to have a feeling of 3DNA has to offer. Do not hesitate to ask if you have any questions.

Xiang-Jun

1424

MD simulations / Re: Ruby scripts for the analysis of MD simulation trajector

« on: March 05, 2011, 07:23:40 pm »

Hi Alpay,

As always, thanks for your feedback and suggestions. Now specifically,

After I submitted my reply, I already changed the line to have the snapshots as numbers only in your script.  Still, I think it is better to have the numbers as default. May be in your next version?

Suggestion taken -- now in v0.3, the first column is just the snapshot number, without the [mono:2688raft]model_[/mono:2688raft] prefix.

By distribution calculation I meant in addition to the means, standard deviations you already supply, it would be nice to have BI/BII distributions averaged over the trajectory, histograms of each parameter (per base pair/base pair step and over the whole structure). I know there is no perfect analysis script/program out there, but the more the initial analysis does the better for the end user. It also encourages people to use it more!

Just to clarify, the scripts currently do not provide mean/std etc statistics -- they are intended to be calculated by the users, using R/Matlab/Octave/Excel  etc. Regarding BI/BII classification of backbone, histograms of parameters etc, I am putting them in my to-do list.

One more suggestion: In the simulations there are always end effects. A general practice among MD people is to remove the first and last base pairs from the analysis to reduce these effects in the subsequent analysis. It would be great to have this option available as a choice in the script.

Suggestion taken -- now v0.3 has a new [mono:2688raft][red:2688raft]-e[/red:2688raft][/mono:2688raft] option to accomplish just what you asked. Moreover, the two ends can be asymmetrical, e.g., [mono:2688raft][red:2688raft]-e 1 2[/red:2688raft][/mono:2688raft] to remove the first and the last two bps from the extracted parameter list.

Download v0.3 to have a try, and report back any issues you experience; and of course, any new suggestions!

Xiang-Jun

1425

MD simulations / Re: average values from MD simulations

« on: February 28, 2011, 12:13:19 am »

Hi Ara,

Thanks for using 3DNA, and for posting your question(s) in the forum. I am glad to hear that you found the Ruby scripts useful to your analysis of drug-DNA MD simulations -- at the very least, I take it as yet another user confirmation that the Ruby scripts are working as expected.

I am using the program to analyze drug-DNA MD simulations and would like to plot base pairs vs Twist (or Rise or Roll) values and therefore I would like to find an average of those values for each base pair over the simulation. Does bp_helical or bp_step provide that kind of info? Or I will need to write my own script for that?

Again, since I have no direct MD-simulation experience, my understanding could be incomplete. If I guess it correctly, your MD simulations should include many snapshots. The output file for each parameter (e.g.,[mono:1b7xkmyt]x3dna_md_roll.out[/mono:1b7xkmyt] for Roll) from the Ruby scripts, [mono:1b7xkmyt]x3dna_md.rb/extract_par.rb[/mono:1b7xkmyt], contains tabulated values arranged in a m-by-n matrix, where m is the number of models/snapshots, and n is the number of base-pair steps. As noted in the initial release post, "The output parameter table is intended to be fed into R/Matlab/Octive/Excel etc for statistical analysis or visualization." Specially, I decided deliberately not to calculate mean/std etc statistics, even though it should be straightforward to add them.

On the other hand, the files "bp_helical.par" and "bp_step.par" are from each run of the [mono:1b7xkmyt][red:1b7xkmyt]analyze[/red:1b7xkmyt][/mono:1b7xkmyt] program on a snapshot, i.e., they are from native 3DNA output, and are overwritten each time unless you renamed them. Put another way, these two files are not related to your MD analysis; instead they are intended to be used with the [mono:1b7xkmyt][red:1b7xkmyt]rebuild[/red:1b7xkmyt][/mono:1b7xkmyt] program to construct DNA structures as specified by the parameters.

HTH,

Xiang-Jun