Companion data and code

This repository contains extra information and code for the article Passive and active colloidal chemotaxis in a microfluidic channel: mesoscopic and stochastic models by Pierre de Buyl and Laurens Deprez. The preprint is available at [arXiv:1701.05020] and will be available also in the ORCID profile of Pierre de Buyl.

The license for the code in this repository is the 3-clause BSD, found in the file LICENSE.

Code and platform information for the mesoscopic simulations

The mesoscopic simulations are performed with the code RMPCDMD that is implemented in Fortran 2008 with a C extension.

Hardware

The simulations were run at the supercomputer ThinKing operated jointly by the KU Leuven and Hasselt University.

The hardware specifications are summarized with the following command-line programs:

user@hpc-p-login-2:~$ uname -mosv
Linux #1 SMP Mon Oct 24 10:22:33 EDT 2016 x86_64 GNU/Linux
user@hpc-p-login-2:~$ grep -E '(^model name|^vendor_id|^flags)' /proc/cpuinfo  | sort | uniq
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good xtopology nonstop_tsc aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm ida arat epb xsaveopt pln pts dtherm tpr_shadow vnmi flexpriority ept vpid fsgsbase smep erms
model name      : Intel(R) Xeon(R) CPU E5-2680 v2 @ 2.80GHz
vendor_id       : GenuineIntel

Compilers

The compilers used are, for C:

user@hpc-p-login-2:~$ gcc --version
gcc (GCC) 4.9.2
Copyright (C) 2014 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

with flags -Wall -std=c99 -O3.

For Fortran:

user@hpc-p-login-2:~$ gfortran --version
GNU Fortran (GCC) 4.9.2
Copyright (C) 2014 Free Software Foundation, Inc.

GNU Fortran comes with NO WARRANTY, to the extent permitted by law.
You may redistribute copies of GNU Fortran
under the terms of the GNU General Public License.
For more information about these matters, see the file named COPYING

with flags -fopenmp -Wall -Wextra -Wconversion -std=f2008-pedantic -O3.

Platform information for the stochastic simulations and data analysis

The stochastic simulations and data analysis were run with Python 3.4 using the interpreter and libraries listed below:

Platform: linux
Python: 3.4.3 (default, Nov 17 2016, 01:08:31)
[GCC 4.8.4]
Machine and architecture x86_64 64bit ELF
NumPy: 1.12.0
SciPy: 0.18.1
Matplotlib: 2.0.0
h5py: 2.2.1

The Python code is provided in Jupyter notebooks.

Information on running and analyzing the simulations

Chemotaxis simulations

To reproduce the simulations of the paper, you need to install the software RMPCDMD. After installation of RMPCDMD, with the rmpcdmd command-line program available in your terminal, and of the Python program sftmpl (pip install --user sftmpl, Python 2 and 3 compatible. This program is used to generate parameter files from a template), you may run the command

make passive_sphere RUN=001 EPS=1.00

and repeat for required the number of realizations. This process should be replicated with the values active_sphere and nanomotor replacing passive_sphere and the values 0.25, 0.50, 2.00 and 4.00 replacing 1.00.

The data from these simulations can be analyzed with the notebook colloidal_chemotaxis.ipynb. The notebook also contains a full implementation of the stochastic model for chemotaxis.

For information, these simulations required about 700 core days.

Diffusion simulations

To reproduce the equilibrium simulations used to calibrate the dimer nanomotor model, run the RMPCDMD program chemotactic_dimer with the command

rmpcdmd run chemotactic_cell dimer_diffusion.p cceq_001.h5 auto

for the required number of realizations.

The data from these simulations can be analyzed with the notebook diffusion.ipynb.

For information, these simulations required about 1600 core days.

Constant gradient model

To reproduce the "constant gradient" stochastic simulations, first compile the Cython module. In the cg_model directory, execute

python3 setup.py build_ext --inplace

then run the simulations 40 times for N_only

python3 run_cg_nm.py N_only_001.h5 --n-steps 100000 --n-inner-steps 25

and 20 times for N_C

python3 run_cg_nm.py N_C_001.h5 --n-steps 100000 --n-inner-steps 25 --C-force