MC DDF, a Computer Program for Generating Distance Distribution Functions from Model Particles using Monte Carlo Chord Sampling

Monte Carlo Distance Distribution Functions (MC_DDF) is a computer program for the generation of distance distribution functions from three-dimensional (3D) models using Monte Carlo distance distribution chord sampling. The technique avoids the problem that exists with finite-element techniques of the need to consider the size of the smallest structure of interest. The program functions by selecting a representative selection from the set of all possible samples within the model, gradually building up a statistically averaged distance distribution function. It is shown that, by a suitable choice of class width (a user parameter), the estimator can be made asymptotically unbiased such that the bias and variance approach zero as the number of chord samples increases. The program has an interface to a 3D modelling package that allows the design and visualization of model particles prior to generating distance distributions. Input can also be accepted from a text data file. Results are produced where the output from the method is compared with analytical model functions. The use of the 3D modelling package is demonstrated using a simple sphere-chain model and a more complex chromatin model. It is shown that a good approximation to the model distance distribution function is obtainable in a relatively short time on a modern PC.