My question is about the meaningful number of particles in the simulation. So, is it a “must” to have very high number of hits in “beamOn” command, to be able to call the results reliable? I understand that statistics plays in, but my question comes from what I explore:
I am running the “dnadamage1” example, in which I am counting the DNA strand breaks with respect to the deposited energy. I reach to the maximum number of breaks (it is a limit in my application) at doses collected with the “beamOn=100”. So, I get the relevant results with “beamOn=100”, and above that, everything is ruined.
Now, is my simulation unreliable or the results are inaccurate because I run it with 100 hits?
I’m not sure about your application, but if you are counting the number of impinging particles it takes to break all of the bonds, then I would set this up to run many independent simulations. Then look at the distribution of the counts it takes to reach that point. That could tell you something interesting. Hopefully that informs what information you’re after.
I am not sure why the simulation results are “unreliable” with a number of incident particles above 100. Maybe try to calculate the same physical quantity per incident particle?
Indeed I have done such an assessment and I know the approximate number of fired particles (or the equivalent deposited energy) that makes all DNA wires to break in the bundle (I have a bundle of linear DNA molecules that I need to have a few of them unbroken after irradiation). Of course, in my application, having a break in a DNA does not correspond to breaking of all bonds.
My main concern was about the outcome of the assessment that I just mentioned; that the number of event/particles that is desired for my purpose (say having 10 unbroken wires out of 36 in the bundle), might not be statistically rich enough to call the results reliable.
So, imagine that after 100 alpha irradiation, I have 10 unbroken wires. I thought that from the view point of MC calculations, 100 events might too low “statistically speaking”. But, later I found 10.3390/cancers13194940] in which a “dose-based” irradiation has been presented. In this study the criterion is “reaching to the dosage of 1 Gy”. So, based on the LET of particles, irradiation of different number of particles is required to reach irradiation of 1 Gy .
Therefore, my conclusion is that if you reporting number of impinging particles based on a certain criterion, having low number of events would not make a problem. Of course, proper number of repetitions is needed to obtain the uncertainty of the results.
Sorry, I guess I have not presented a clear description of the problem.
I have created a bundle of linear double strand DNA fragments (lets call DNA them wires). Since I am after the electrical response of the bundle, I need to have some of the wires unbroken upon irradiation (no unbroken wire means no current!). Now, I obtain the desired number of unbroken wires through a certain amount of deposited energy (dose) or 100 events. If I do the calculations with >100 events, all of the wires would be broken. Of course you might suggest to increase the number of wires or their length, but our computational capability is a limit here.
Now, the question that I raise is as follows (re-stating part of my answer to @Stephen_Coleman):
Imagine that after 100 alpha irradiation, I have 10 unbroken wires. I thought that from the view point of MC calculations, 100 events might too low “statistically speaking”. But, later I found 10.3390/cancers13194940] in which a “dose-based” irradiation has been presented. In this study the criterion is “reaching to the dosage of 1 Gy”. So, based on the LET of particles, irradiation of different number of particles is required to reach irradiation of 1 Gy .
Therefore, my conclusion is that if you reporting number of impinging particles based on a certain criterion, having low number of events would not make a problem. Of course, proper number of repetitions is needed to obtain the uncertainty of the results.
I would appreciate if you share your thoughts on this.