Activity for oxDNA
-
oxDNA released /v3.7.0/v3.7.0 source code.zip
-
-
-
-
syc posted a comment on discussion General Discussion
OK. Got it. Thanks so much.
-
Lorenzo Rovigatti posted a comment on discussion General Discussion
Then that version does not support this syntax. You'll have to add a couple of braces for each particle.
-
The version released in 2019
-
What version are you running?
-
But when I run simulations with this syntax. The information printed out says that the force was only added to the first atom of the particles listed there.
-
Yes, you can use particle = 1-10 to include particles with indexes between 1 and 10, or particle = 1,2,3-5 to include particles with ids 1, 2, 3 4, or any variation thereof. By the way, note that this repository is being discontinued.
-
I want to add repulsive plane to half of the particles in the structure. Is there a way of adding the force in one pair of bracket rather than adding the force to each particle specifically?
-
-
-
-
Hi! First of all, note that this repo is being discontinued in favour of the github one, so I advise you to post future questions over there. As for the matter at hand, simulations of 50bp systems run much faster on the CPU than on the GPU, and in fact for these sizes it's often best to run VMMC (rather than MD) simulations, which can be done on CPU only. However, we have also explored the possibility of running many small simulations in parallel on the GPU. You can find some results here. However,...
-
Jackson Anderson posted a comment on discussion General Discussion
Hi Lorenzo, This is very helpful. Thanks for the response! Best, -Jackson
-
Hi! First of all, note that this repo is being discontinued in favour of the github one, so I advise you to post future questions over there. As for the matter at hand, simulations of 50bp systems run much faster on the CPU than on the GPU, and in fact for these sizes it's often best to run VMMC (rather than MD) simulations, which can be done on CPU only. However, we have also explored the possibility of running many small simulations in parallel on the GPU. You can find some results here. However,...
-
Hello, My lab is considering getting a computer for running oxDNA simulations locally. We are working with smaller DNA molecules, around 50 bp. Ideally, we'd be able to run a few simulations a day. What hardware would we realistically need for this? Would a normal consumer GPU suffice? Thanks for your help.
-
-
-
-
Bruce E. Camber modified a comment on discussion General Discussion
Can the gap of 7.3561+ degrees with five-tetrahedrons (each sharing an edge with another and all sharing just one edge) be re-created with any of the construction sets like Zometool or using interactive geometry software (IGS) and their dynamic geometry environments (DGEs)? The Zometool can not display a gap.
-
Are there any geometers in the forum? Can Python Tkinter be used to recreate that gap? Are geometries part of condensed matter physics? ...computational physics? ...interpolation & model fitting? ...computational geometry? Is Aristotle's mistake -- not being aware of that gap -- haunting us? Jeffrey C. Lagarias and Chuanming Zong reported about it in the 2012 AMS bulletin, Mysteries in Packing Regular Tetrahedra, Notices of the AMS, Volume 59, Number 11 (1540). In 2015 they received the AMS Conant...
-
I'm sorry but I have no idea what you are talking about. I think you posted in the wrong forum
-
Can the gap of 7.3561 degrees with five-tetrahedrons (each sharing an edge with another and all sharing just one edge) be re-created with any of the construction sets like Zometool or using interactive geometry software (IGS) and their dynamic geometry environments (DGEs)? The Zometool can not display a gap.
-
Hi! Note that this forum is being discontinued in favour of the new repo In principle there are solutions to limit memory usage, but it's hard to say without knowing what you are trying to do. The simplest way would be to reduce the box size you are using,
-
Bingwang Zhang posted a comment on discussion General Discussion
Hi everyone, I'm currently performing pulling simulations with oxDNA, using string as a force and rate = 5e-9. As I understand it, for string, force = ratetimesteps. However, by calculating the unfolded force, the simulation results are much larger than the experimental results. I speculate that it may be that my system is not in a stable state. My question is, is there any way to tell if my system is in a steady state? My system is small, less than a hundred nucleotides. Any suggestions are wel...
-
Hi everyone, I'm currently performing pulling simulations with oxDNA, using string as a force and rate = 5e-9. As I understand it, for string, force = ratetimesteps. However, by calculating the unfolded force, the simulation results are much larger than the experimental results. I speculate that it may be that my system is not in a stable state. My question is, is there any way to tell if my system is in a steady state? My system is small, less than a hundred nucleotides. Any suggestions are wel...
-
niloufar posted a comment on discussion General Discussion
Hi, I'm running a big DNA origami using oxDNA with cuda, but I get this error related to GPU memory (the screenshot of the error message is attached.) The GPU memory of my system is 12GB, but it sounds like we need about 14 GB GPU memory to run the simulation. Is there any solution for this issue? I would appreciate your help on this matter.
-
-
-
-
-
-
-
Hi! Note that this forum is being discontinued in favour of the new repo Having said that, have you tried following the procedure described here and in the examples/NEW_RELAX_PROCEDURE folder
-
Hi! Note that this forum is being discontinued in favour of the new repo Having said that, have you tried following the procedure described here and in the examples/NEW_RELAX_PROCEDURE folder
-
swapnil baral posted a comment on discussion General Discussion
Hi, I want to force a bond to remain paired during simulation. I tried using mutual trap as follows: { type = mutual_trap particle = 0 ref_particle = 17 stiff = 1. r0 = 1.2 } { type = mutual_trap particle = 17 ref_particle = 0 stiff = 1. r0 = 1.2 } But the bonds break in first few steps during relaxation. What can I do to prevent bond 0 and 17 from breaking ? If I want to do this for multiple bonds, can I put all mutual trap pairs in a single file ?
-
Jaka Socan posted a comment on discussion General Discussion
Thank you! I will refer to the relevant papers and come back if I have some specific question.
-
Hey, I'd say your understanding is correct, but you should refer to the relevant papers where the model is discussed. cheers, Lorenzo
-
Hi Megan, Thanks for your reply. The loading rate I used in my experiment was 1 pN/s and the unfolding force was 56 pN; The loading rate used in simulation is 5e-9 simulation unit/time step(8e7 pN/s), and the unfloding force is 90 pN, which is very different. Is it related to the type of force applied (such as string or harmonic trap)? What should I do in this situation? Do you have any suggestions? Thank you. Bing wang
-
Hi Megan, Thanks for your reply. The loading rate I used in my experiment was 1 pN/s and the unfolding force was 56 pN; The loading rate used in simulation is 5e-9 simulation unit/time step(8e7 pN/s), and the unfloding force is 90 pN, which is very different. Is it related to the type of force applied (such as string and harmonic trap)? What should I do in this situation? Do you have any suggestions? Thank you. Bing wang
-
Hi Megan, Thanks for your reply. The loading rate I used in my experiment was 1 pN/s and the unfolding force was 56 pN; The loading rate used in simulation is 5e-9 simulation unit/time step(8e7 pN/s), and the unfloding force is 90 pN, which is very different. What should I do in this situation? Do you have any suggestions? Thank you. Bing wang
-
Megan C Engel posted a comment on discussion General Discussion
Hey Bingwang, The unfolding force for a structure depends on the pulling rate (and how far the system is from equilibrium). For faster experimental pulling rates, we see higher unfolding forces. Because the coarse-grained dynamics in oxDNA are "sped up" compared to experiments (and comparing coarse-grained absolute rates/times to experiments is tricky in any case), it's not surprising that you see higher unfolding forces than in experiment. In effect, your simulations are simply farther from equilibrium...
-
Hi everyone, I'm currently performing pulling simulations with oxDNA, using string as a force and rate = 5e-9. As I understand it, for string, force = ratetimesteps. However, by calculating the unfolded force, the simulation results are much larger than the experimental results. I speculate that it may be that my system is not in a stable state. My question is, is there any way to tell if my system is in a steady state? My system is small, less than a hundred nucleotides. Any suggestions are wel...
-
Hi again, I have so far been using my oxDNA simulations at the room temperature and relative high ionic strength (1 M). I'd now like to discover the temperature and ionic strength dependence of my systems over some reasonable range. When skimming through the relevant literature, I noticed that the force field behaves as intended at least over the range from 290 to 370 K (room to water boiling temperature), while at lower temperatures the stacking interaction tends to become overestimated. Would it...
-
I got my answer Thank you very much
-
You are printing so often that most of the time is spent getting data out of the GPU and to the disk. Try setting print_energy_every and print_conf_interval to larger values (1e4 and 1e6 should be enough).
-
You are printing so often that most of the time is spent getting data out of the GPU and to the disk. Try setting print_energy_every and print_conf_interval to larger numbers (1e4 and 1e6 should work).
-
The interesting point is that the CUDA_EXAMPLE runs with GPU but it does not work for new structure.
-
It seems that the CUDA is active, but the running speed is much slower than the CPU and the system monitor shows that the CPU is 100% active.
-
Hi! Note that this forum (and the whole repo) is being discontinued in favour of the github repo. Your input file looks fine, are you sure the simulation is not running on the GPU?
-
Hi, I have installed the cuda version of oxDNA on my system. When I run an MD sim_type with cuda backend, the program runs without any error, but the GPU does not get involved and only one CPU is involved in my simulation. I have attached my conf and top files along with the input file. Further, I wirte oxDNA <input_file> for running the simulation. I would appreciate your help on this issue.</input_file>
-
Hi! This is the discussion forum for the old oxDNA version, which is being discontinued. Please head over https://github.com/lorenzo-rovigatti/oxDNA and post your issue over there to have a more up-to-date support. Note that your issue may have something to do with your python or oxDNA_analysis_tools version.
-
Abdullah Kazi posted a comment on discussion General Discussion
Hello, I am trying to analyze my trajectory file. I have compiled both oxDNA and oxpy (along with dependencies) and for some reason I can't run the program output_bonds.py as there is a syntax error "def parse_header(e_txt:str) -> List[str]:" Would love any help or alternate ways to analyze the bondings in the trajectory file.
-
Gathering enough statistics can take up to days or weeks depending on the simulation, so I'm not surprised that it's taking its time. Unfortunately 74 nucleotides is too small of a system to benefit from using GPUs.
-
Mengsheng Zhang posted a comment on discussion General Discussion
Hi Lorenzo, I'm running a system with 74 nucleotides of DNA, running on the CPU for a long time, and the program is still running. Does this structure run faster on GPU?
-
The number of steps required will heavily depend on the specific system, as well as on the conditions you are simulations. It is not uncommon having to simulate up to 10^8, 10^9 steps, and maybe average over different runs. My humble advise would be to get some experience in simulating this sort of things, starting with very simple systems (e.g. double strands).
-
Hi, I have modified some of the parameters according to the suggestions you provided. If I want to obtain equilibrium result, then is the larger the number of steps, the better ? The number of steps I currently set is 2e5.
-
If you want to obtain equilibrium results then you have to pull really slowly, and your simulation needs to be really long. How long? That depends on your specific system. Your input files look fine, but I guess that you'll have to be careful in picking your parameters (F0, rate, etc)
-
Hi, Thanks for the reply. I have a question in regards to a project that I am working on. I first got the prova.top and .conf files using oxviewer, then referred to RELAX_INITIAL_CONFIGURATION to relax structure, and then ran MD simulation. I was simulating the force-extension of nucleic acid, requiring one end to be fixed and the other end (not the end) to be applied. However, it seems that the structure obstained at last is not correct (the nucleic acid is still folding and when the force increases,...
-
Hi, Thanks for the reply. I have a question in regards to a project that I am working on. I first got the prova.top and .conf files using oxviewer, then referred to RELAX_INITIAL_CONFIGURATION to relax structure, and then ran MD simulation. I was simulating the force-extension of nucleic acid, requiring one end to be fixed and the other end (not the end) to be applied. However, it seems that the structure obstained at last is not correct (the nucleic acid is still folding). Could you please help...
-
Hi, Thanks for the reply. I have a question in regards to a project that I am working on. I first got the prova.top and .conf files using oxviewer, then referred to RELAX_INITIAL_CONFIGURATION to relax structure, and then ran MD simulation. I was simulating the force-extension of nucleic acid, requiring one end to be fixed and the other end (not the end) to be applied. However, it seems that the structure obstained at last is not correct (the nucleic acid is still folding). Could you please help...
-
I think I can see some stretched bonds in the "relaxed" configuration. You may want to try relaxing it a bit more. You could also try to relax it on GPU by using the langevin thermostat with a tighter coupling, plus a smaller dt.
-
Hi! Yes, you can do it with oxDNA. You can have a look at the examples/TRAPS folder (using the the new repo on github).
-
Pranav Sharma modified a comment on discussion General Discussion
Hello, I am simulating an RNA tetrahedron origami (designed using pyDAEDALUS, another origami design software) at 37 degrees Celsius using RNA2 force field, CUDA backend, and Langevin thermostat. I have first done (without issue) 1e6 MC steps of MC relaxation and 3e6 steps of MD relaxation on CPU backend using RNA_Relax force field. At the end I get what looks like a decently relaxed structure (see relaxed_tetrahedron.png image attached). When I go to simulate this using the RNA2 force field on CUDA,...
-
Hello, I am simulating an RNA tetrahedron origami (designed using pyDAEDALUS, another origami design software) at 37 degrees Celsius using RNA2 force field, CUDA backend, and Langevin thermostat. I have first done without issue 1e6 MC steps of MC relaxation and 3e6 steps of MD relaxation on CPU backend using RNA_Relax force field. At the end I get what looks like a decently relaxed structure (see relaxed_tetrahedron.png image attached). When I go to simulate this using the RNA2 force field on CUDA,...
-
Hello, I am simulating an RNA tetrahedron origami (designed using pyDAEDALUS, another origami design software) at 37 degrees Celsius using RNA2 force field, CUDA backend, and Langevin thermostat. I have first done without issue 1e6 MC steps of MC relaxation and 3e6 steps of MD relaxation on CPU backend using RNA_Relax force field. At the end I get what looks like a decently relaxed structure (see relaxed_tetrahedron.png image attached). When I go to simulate this using the RNA2 force field on CUDA,...
-
Hi, I am new user for oxDNA and I recently want to simulate the DNA hairpin unfolding force(the force-extension curve in the attachment), In it, one end of DNA hairin is fixed and the other end is forced. Is there any way to solve this problem? Or how can I simulate it using oxDNA? Thanks! Best, Mengsheng Zhang
-
Hi, I am new user for oxDNA and I recently want to simulate the DNA hairpin unfolding force(the force-extension curve in the attachment), Is there any way to solve this problem? Or how can I simulate it using oxDNA? Thanks! Best, Mengsheng Zhang
-
Hi, thanks, this seems to do the trick!
-
VMMC starts having issues when a cluster can interact with multiple copies of the same particles, so that may be why you are experiencing issues. In general I would always use a box with sides that are larger than the contour length of the molecule (unless you use external forces that force the molecule to take an orientation that is parallel to the long side of a non-cubic box). As for useful settings, you may want to play with the maxclust key, which sets the maximum number of particles to be moved...
-
Hi, I have somewhat related issue, I'm now doing a single long dsDNA chain simulation. My idea was to decrease the box size in order to increase the concentration. However, I'm now noticing strange artifacts start to appear in VMMC with smaller boxes (molecule gets shredded to fragments after some number of steps). Is there perhaps a rule of a thumb regarding how small a box size can still be used to simulate a molecule with certain contour length? Alternatively, are there any VMMC setting available...
-
Hi! I think there's something wrong with the topology: there are no hydrogen bonds in the initial configuration. For historical reasons the oxDNA topology should be specified in the 3'->5' direction, while the rest of the (sane) world uses 5'->3'. Maybe this is the issue?
-
Hi, I am having difficulties in relaxing the initial DNA origami structure which I created using ATHENA. The minimization step goes well, but for the relaxation, after a few steps the structure falls apart. I tried to change some parameters such as dt, thermostat, temperature... but unfortunately none of them worked. I have attached my conf, top, and input files.
-
Arjav Shah posted a comment on discussion General Discussion
Thanks!
-
No, there is no such external force at the moment.
-
Hello, I had a follow-up question based on the previous discussion - In oxDNA2, is it possible to apply an external force on a specified particle that is dependent on the position of the particle wrt center of mass of the object it is a part of? Basically, I want to apply a force that needs to be adjusted at every time step based on the spatial coordinates of a bead wrt center of mass. Thank you!
-
-
-
-
Signal 15 (SIGTERM) is usually issued by some other process to kill the program (see here for example).
-
I am encountering following error a first time. What does it mean? INFO: seeding the RNG with 101 INFO: Initializing backend INFO: Backend precision: float INFO: Simulation type: MD INFO: The generator will try to take into account bonded interactions by choosing distances between bonded neighbours no larger than 2.000000 INFO: Using '' as the input for sequence-dependent values INFO: Running Debye-Huckel at salt concentration = 0.5 INFO: Using different widths for major and minor grooves INFO: Setting...
-
The answer depends on your system's details. Decreasing the system size effective increases the DNA concentration. However, if you are studying a structure where all strands never detach then this should be important. In this case you should be careful that the box size is large enough that the structure doesn't feel itself through periodic boundary conditions. About your other question, the GPU code's performance decreases as the box size increases, but the actual scaling depends on the specific...
-
Hi, I have a very basic question about choosing the box size in oxDNA. Is it okay to decrease the box size as long as everything is inside the box when viewed in oxview. How does the simulation run time scale as the function of box size for the given DNA ? What is the best way to decide the box dimensions ? Thank you.
-
That is helpful! Thank you so much, Lorenzo.
-
The pressure is defined as the trace of the virial stress tensor of the system, and therefore the osmotic pressure refers to DNA and DNA-DNA interactions only, since the effect of salt is taken into account implicitly. Note that the partial osmotic pressure of DNA systems (simulated at realistic concentrations) is so small that I'm not sure you would be able to sample it accurately enough to compare it to "real" values.
-
Yes, you can use that one and set particle = all so that it applies to all the particles (or use a subrange like particle = 1-100 to apply it to a subset of the system.
-
Yes, exactly right. Using 'string' might be able to do it?
-
So you need a way of pulling all particles towards a particular direction? There are external forces to do that.
-
Hi Lorenzo, Thanks for your reply! I understand, I will try coding it up. However, is it possible to apply a constant electric field (constant in time and space) inside the simulation box?
-
Hi, nope, there is no such external forces as of now. You could code it up, but you would need some deep-ish knowledge of the codebase in order to do that.
-
Hi, thank you for explanation and help. Decreasing the pressure quantity stabilized the system which is now fluctuating normally. I presume one could calculate osmotic pressure of the system quite accurately, as most of the solute in the system is represented by salt (at default concentration of 1.0). Is that logic correct or do I actually need to only estimate the partial osmotic pressure of DNA/RNA molecules? Could you maybe also explain to me what does the quantity labeled "backend_info", coming...
-
Hello, I am looking to simulate the effect of an external electric field on a DNA strand. Are there ways to apply (constant in time but spatially-distributed) external field in the box and not on any given particles? I understand that the current force options require us to specify particles (particle (int) option) which makes it harder to apply fields that are spatially dependent. In short, can I apply forces that are space-dependent and not on a particular particle? Thanks in advance for your ...
-
Hello, I am looking to simulate the effect of an external electric field on a DNA strand. Are there ways to apply (constant but spatially-distributed) external field in the box and not on any given particles? I understand that the current force options require us to specify particles (particle (int) option) which makes it harder to apply fields that are spatially dependent. In short, can I apply forces that are space-dependent and not on a particular particle? Thanks in advance for your help!
-
Since oxDNA is an implicit solvent model, the pressure you set (and that you can measure in the simulation) is the osmotic pressure of the solution, and not the overall pressure, which is general much higher than the former.
-
Thank you! I tried to estimate the room pressure in reduced units used in oxDNA. As for MC - I have tried that and it behaved similarly (shrinking volume, increasing pressure). Note that my system is rather large thus I can't get it evolved much with MC even if I use step optimization. I will try with lower numbers and come back in case I get an improvement. Cheers!
-
Hi, I need to use NPT ensemble for my simulations, which is why I'm trying to run an NPT equilibration of a system (10000 base pairs, 0.1 μm3 box) already equilibrated under NVT conditions using the following instructions: #### PROGRAM PARAMETERS #### # Only DNA_relax works with this system interaction_type = DNA2 sim_type = MD backend = CUDA backend_precision = mixed debug = 1 seed = 16 max_density_multiplier = 4 #### SIM PARAMETERS #### ensemble = NPT list_type = cells #NPT req steps = 1e7 newtonian_steps...
-
Hi, I need to use NPT ensemble for my simulations, which is why I'm trying to run an NPT equilibration of a system (10000 base pairs, 0.1 μm3 box) already equilibrated under NVT conditions using the following instructions: #### PROGRAM PARAMETERS #### # Only DNA_relax works with this system interaction_type = DNA2 sim_type = MD backend = CUDA backend_precision = mixed debug = 1 seed = 16 max_density_multiplier = 4 #### SIM PARAMETERS #### ensemble = NPT list_type = cells #NPT req steps = 1e7 newtonian_steps...
-
It's very hard to tell what's happening without having the whole trajectory file. Can you attach a folder containing files that can be used to reproduce the error?
1 >
