Sampling of 3 variables

This dataset consists of n=500 isometric twitches simulated with 1 ms resolution.

Simulations were run with values of the following variables selected based on a Latin hypercube design.

Variable Description Min value (multiple of base) Max value (multiple of base) Base value
m_1_2_2_1 A myosin rate constant 10-0.5 100.5 200
m_1_3_2_1 A myosin rate constant 10-0.5 100.5 350
Ca_k_act A rate constant for Ca2+ handling 10-0.5 100.5 350

In this example, m_1_2_2_1 ranged between 10-0.5 * 200 = 63 and 100.5 * 200 = 632.

How to generate these data

  • Follow the instructions for the FiberSim demo but …
  • … run the command python FiberPy.py sample "<path_to_this_repo>/simulations/n_vars_3/base/setup.json

Example of a single trial

Single trial

Pair-plot

This figure shows the distribution of the parameter values. The histogram has a concave profile because the rate constant was sampled in log space and drawn in linear space.

Pair plot

500 trials superposed

Single trial

Note that for some parameter values, the muscle activated at low Ca2+ so that force rose early in the simulation and did not relax thereafter.

Data

To keep files of manageable width, the 500 simulation traces are divided into 5 files, each containing records for 100 traces.

Each file is a 2500 x 301 tab-delimited text file with columns arranged as follows:

  • Time (s)
  • Ca2+ conc (M) for simulation n
  • Half-sarcomere length (nm) for simulation n
  • Stress (N m-2) for simulation n
  • Ca2+ conc (M) for simulation n+1
  • Half-sarcomere length (nm) for simulation n+1
  • Stress (N m-2) for simulation n+1
  • Ca2+ conc (M) for simulation n+1
  • Half-sarcomere length (nm) for simulation n+1
  • Stress (N m-2) for simulation n+1
  • …
  • Ca2+ conc (M) for simulation n+99
  • Half-sarcomere length (nm) for simulation n+99
  • Stress (N m-2) for simulation n+99

The links are:

Parameter values is an Excel file where the nth row contains the parameter values for simulation n