Methods, systems, and computer readable media for data analysis and inference of particle diffusion in target materials and target material simulants

1. A method for data analysis and inference of particle diffusion in mucus barriers and permeable biomaterials, the method comprising:

collecting experimental data of observed movements of particles through samples of a target material, wherein collecting the experimental data includes using video microscopy and particle tracking software to capture changes in particle positions over time in the samples, wherein the samples comprise mucus or permeable biomaterial samples;

grouping the particles into a number of distinct clusters based on the observed movements of particles;

analyzing the changes in particle positions over time to determine, for each cluster, a stochastic diffusive process that is being observed for particular particles in the cluster and parameters of the stochastic diffusive process, wherein the parameters of the stochastic diffusive process of each cluster are sampled from a Gaussian distribution that is determined, using an expectation maximization algorithm, to have generated the cluster;

using the stochastic diffusive processes for the clusters to simulate diffusion of particles through the target material, wherein simulating the diffusion of particles through the target material comprises performing a plurality of simulations, wherein each simulation simulates movement of particles through a target material of a particular thickness, and includes, for each particle simulated:

using parameters of the Gaussian distribution of the cluster to which the particle is assigned to determine standard deviations from the Gaussian distribution;

converting the standard deviations into Hurst parameter and pre-factor values;

using the Hurst parameter and pre-factor values to simulate fractional Brownian motion for the particle, which results in a simulated path of the particle through the target material: and

using the plurality of simulated paths to calculate passage time distributions through the target material of the particular thickness, wherein data from the plurality of simulations is used to calculate passage time distributions for each of a plurality of thicknesses of the target material;

using results from the simulating to model passage times of the particles in the target material as a function of layer thickness of the target material; and

predicting, using the modeling of the passage times determined from the results from the simulating, uptake of a drug or nanoparticles in the target material prior to a start of animal or clinical drug trials involving administering the nanoparticles or the drug to subjects through the target material.

2. The method of claim 1 wherein collecting experimental data comprises infusing the samples with particles of interest and measuring positions of the particles at discrete time intervals.

3. The method of claim 1 wherein analyzing the collected data comprises assessing heterogeneity of the target material with respect to particle size.

4. The method of claim 1 wherein the clusters comprise distinct populations of particles that differ from populations of particles in other clusters based on particle movement step size distribution.

5. The method of claim 4 wherein grouping the particles into a number of distinct clusters comprises calculating standard deviations of step size distributions for each particle, identifying populations of particles with similar standard deviations, and assigning particles with similar standard deviations to one of the clusters.

6. The method of claim 4 wherein grouping the particles into distinct clusters comprises applying a Gaussian mixture model where a number of Gaussian distributions is equal to the number of distinct clusters.

7. The method of claim 6 wherein the expectation maximization algorithm comprises an iterative expectation maximization algorithm.

8. The method of claim 1 wherein using results from the simulating further comprises using simulation data to determine clinically relevant factors, make predictions regarding predicted behavior of experimental data, and determine how well the predicted behavior fits an actual behavior observed in the experimental data.

9. The method of claim 1 comprising verifying the results from the simulating against at least one of:

the experimental data; and

data collected during a subsequent validation experiment.

10. The method of claim 1 wherein collecting experimental data of observed movements of particles through samples of a target material includes collecting experimental data of movements of particles of different diameters though the target material.

11. A system for data analysis and inference of particle diffusion in mucus barriers and permeable biomaterials, the system comprising:

a data storage device for storing collected experimental data of observed movements of particles through samples of a target material, wherein the collected experimental data includes data generated by particle tracking software through analysis of video microscopy data and indicating changes in particle positions over time in the samples, wherein the samples comprise mucus or permeable biomaterial samples; and

a hardware processor for grouping the particles into a number of distinct clusters based on the observed movements of particles, analyzing the collected experimental data to determine, for each cluster, a stochastic diffusive process that describes the observed particle movement in the mucus or permeable biomaterial of particles in a cluster and parameters of the stochastic diffusive process, wherein the parameters of the stochastic diffusive process of each cluster are sampled from a Gaussian distribution that is determined, using an expectation maximization algorithm, to have generated the cluster, wherein the hardware processor is further configured for using the stochastic diffusive processes for the clusters to simulate diffusion of particles through the target material, using results from the simulating to model passage time distributions of the particles in the target material as a function of layer thickness of the target material and predicting, using the modeling of the passage time distributions determined from the results from the simulating, uptake of a drug or nanoparticles in the target material prior to a start of animal or clinical drug trials involving administering the nanoparticles or the drug to subjects through the target material, wherein simulating the diffusion of particles through the target material comprises performing a plurality of simulations, wherein each simulation simulates movement of particles through a target material of a particular thickness, and includes, for each particle simulated:

using parameters of the Gaussian distribution of the cluster to which the particle is assigned to determine standard deviations from the Gaussian distribution;

converting the standard deviations into Hurst parameter values;

using the Hurst parameter and pre-factor values to simulate fractional Brownian motion for the particle, which results in a simulated path of the particle through the target material; and

using the plurality of simulated paths to calculate passage time distributions through the target material of the particular thickness, wherein data from the plurality of simulations is used to calculate passage time distributions for each of a plurality of thicknesses of the target material.

12. The system of claim 11 wherein the collected experimental data was collected by infusing the samples with particles of interest and measuring positions of the particles at discrete time intervals.

13. The system of claim 11 wherein analyzing the collected data comprises assessing heterogeneity of the target material with respect to particle size.

14. The system of claim 11 wherein the clusters comprise distinct populations of particles that differ from populations of particles in other clusters based on particle movement step size distribution.

15. The system of claim 14 wherein grouping the data into distinct populations based on step size distribution comprises calculating standard deviations of step size distributions for each particle, identifying populations of particles with similar standard deviations, and assigning particles with similar standard deviations to one of the clusters of particles.

16. The system of claim 14 wherein grouping the particles into clusters comprises applying a Gaussian mixture model where a number of Gaussian distributions is equal to the number of distinct clusters.

17. The system of claim 16 wherein the expectation maximization algorithm comprises an iterative expectation maximization algorithm.

18. The system of claim 11 wherein using results from the simulating further comprises using simulation data to determine clinically relevant factors, make predictions regarding predicted behavior of experimental data, and determine how well the predicted behavior fits an actual behavior observed in the experimental data.

19. The system of claim 11 comprising verifying the results from the simulating against at least one of:

the experimental data; and

data collected during a subsequent validation experiment.

20. The system of claim 11 wherein the collected experimental data of observed movements of particles through samples of a target material includes experimental data of movements of particles of different diameters though the target material.

21. A non-transitory computer readable medium having stored thereon executable instructions that when executed by a processor of a computer control the computer to perform steps comprising:

collecting experimental data of observed movements of particles through samples of a target material, wherein collecting the experimental data includes using video microscopy and particle tracking software to capture changes in particle positions over time in the samples, wherein the samples comprise mucus or permeable biomaterial samples;

grouping the particles into distinct clusters based on the observed movements of particles;

analyzing the changes in particle positions over time to determine, for each cluster, a stochastic diffusive process that approximates particle statistics for particles in the cluster and parameters of the stochastic diffusive process, wherein the parameters of the stochastic diffusive process of each cluster are sampled from a Gaussian distribution that is determined, using an expectation maximization algorithm, to have generated the cluster;

using the stochastic diffusive processes for the clusters to simulate diffusion of particles through the target material, wherein simulating the diffusion of particles through the target material comprises performing a plurality of simulations, wherein each simulation simulates movement of particles through a target material of a particular thickness, and includes, for each particle simulated:

using parameters of the Gaussian distribution of the cluster to which the particle is assigned to determine standard deviations from the Gaussian distribution;

converting the standard deviations into Hurst parameter and pre-factor values;

using the Hurst parameter and pre-factor values to simulate fractional Brownian motion for the particle, which results in a simulated path of the particle through the target material; and

using the plurality of simulated paths to calculate passage time distributions through the target material of the particular thickness, wherein data from the plurality of simulations is used to calculate passage time distributions for each of a plurality of thicknesses of the target material;

using results from the simulating to model passage time distributions of the particles in the target material as a function of layer thickness of the target material; and

predicting, using the modeling of the passage time distributions determined from the results from the simulating, uptake of a drug or nanoparticles in the target material prior to a start of animal or clinical drug trials involving administering the nanoparticles or the drug to subjects through the target material.

22. The non-transitory computer readable medium of claim 21 wherein collecting experimental data of observed movements of particles through samples of a target material includes collecting experimental data of movements of particles of different diameters though the target material.