A model of High Speed Atomic Force Microscopy (HS-AFM), based on statistics from imp's Nuclear Pore Complex transport module.
Tested on the following systems:
- Windows 11 Version 10.0.22621 Build 22621
- Debian GNU/Linux 11 (bullseye)
- Python 3.6 or higher
- packages listed in requirements.txt
- Clone the repository:
git clone https://github.com/xroi/HS-AFM-Simulation.git
- Install the required packages:
pip install -r requirements.txt
The typical installation time should be a few minutes.
This example is demonstrated using the demo dataset, available in the demo_dataset folder.
The input for the simulation is a sequence of .hdf5 files containing spatial density statistics at increasing
timepoints.
These can be generated using the npctransport module in imp.
To run the simulation, use the following command:
python src/main.py @demo_dataset/args.txt
args.txt is a file containing the following arguments:
-h, --help show this help message and exit
--npc-simulation, --no-npc-simulation
In the case of --npc-simulation flag, the npc
simulations run 'live'. In the case of --no-npc-
simulation flag, the program uses files in the folder
specified with --existing_files_path
--input-path INPUT_PATH
Path to the folder containing hdf5 density map files,
to be used in the case of --no-npc-simulation flag.
--input-suffix INPUT_SUFFIX
Files should be in input-path and named
<delta_time_in_ns>.<input-suffix>
--read-from-gzip, --no-read-from-gzip
If true, allows reading from gzip compressed HDF5
files.
--simulation-start-time-ns SIMULATION_START_TIME_NS
Start time of the AFM simulation in nanoseconds. (NPC
will be simulated since 0)
--simulation-end-time-ns SIMULATION_END_TIME_NS
End time of the AFM simulation in nanoseconds. (NPC
will be simulated since 0)
--interval-ns INTERVAL_NS
Interval time of the NPC simulation in nanoseconds.
--voxel-size-a VOXEL_SIZE_A
Size of each voxel in the NPC simulation in Angstrom.
(Should be the same as specified inthe npc simulation
configuration file).
--tunnel-radius-a TUNNEL_RADIUS_A
Radius of the NPC tunnel in Angstrom. (Should be the
same as specified inthe npc simulation configuration
file).
--slab-thickness-a SLAB_THICKNESS_A
Thickness of the NPC slab in angstrom. (Should be the
same as specified inthe npc simulation configuration
file).
--statistics-interval-ns STATISTICS_INTERVAL_NS
(Should be the same as specified in the npc simulation
configuration file).
--torus-slab, --no-torus-slab
Specifies if the slab is a torus. should be the same
as in the original simulation.
--separate-n-c, --no-separate-n-c
Specifies if fg chains are seperated into N and C
parts.
--min-x-coord MIN_X_COORD
Specifies the first pixel on the X axis on which the
simulation is ran (inclusive). Count starting from 0.
--max-x-coord MAX_X_COORD
Specifies the last pixel on the X axis on which the
simulation is ran (not inclusive). Count starting from
0.
--min-y-coord MIN_Y_COORD
Specifies the first pixel on the Y axis on which the
simulation is ran (inclusive). Count starting from 0.
--max-y-coord MAX_Y_COORD
Specifies the last pixel on the Y axis on which the
simulation is ran (not inclusive). Count starting from
0.
--min-z-coord MIN_Z_COORD
Specifies the first pixel on the Z axis on which the
simulation is ran (inclusive). Count starting from 0.
--max-z-coord MAX_Z_COORD
Specifies the last pixel on the Z axis on which the
simulation is ran (not inclusive). Count starting from
0.
--vertical-scanning, --no-vertical-scanning
If true, scans the lines vertically.
--z-func {z_test,z_test2}
--calc-needle-threshold, --no-calc-needle-threshold
If true, calculates the threshold using the
median_threshold function in utils.py.
--calc-threshold-r-px CALC_THRESHOLD_R_PX
The radius around the center to get the median value
from in pixels. (for calculating the threshold if
calc_needle_threshold is true)
--calc-threshold-frac CALC_THRESHOLD_FRAC
The fraction of the median to take for the threshold.
(for calculating the threshold if
calc_needle_threshold is true)
--tip-custom-threshold TIP_CUSTOM_THRESHOLD
The density under which the needle ignores. Used for
all z funcs.
--tip-radius-px TIP_RADIUS_PX
Determines how far around the origin pixel the needle
considers for determining pixel height. (Assuming ball
shape). Should be greater than 1. Only used for z_top
z func.
--floaters-resistance, --no-floaters-resistance
Determines whether floaters (NTRs and Passive
Molecules) offer resistance to the simulated tip.
--fgs-sigma-a FGS_SIGMA_A
The sigma value for the normal distribution used to
weigh fg chains.
--floaters-sigma-a FLOATERS_SIGMA_A
The sigma value for the normal distribution used to
weigh floaters.
--floater-size-factor FLOATER_SIZE_FACTOR
Determines the factor by which the floater size in
angstrom is multiplied which is then added to the
floater weight. (0.125 to be proportional with 8A fg
beads)
--floater-distribution-factor FLOATER_DISTRIBUTION_FACTOR
Determines the factor by which the distribution is
multiplied which is then added to the floater weight.
--floater-general-factor FLOATER_GENERAL_FACTOR
Determines the factor by which the final weight of the
floater is multiplied
--time-per-line-ns TIME_PER_LINE_NS
Determines the amount of time it takes for a needle to
pass a full line. Mutually exclusive with time-per-
pixel-ns.
--time-per-pixel-ns TIME_PER_PIXEL_NS
Determines the amount of time it takes for a needle to
pass a single pixel. Mutually exclusive with time-per-
line-ns.
--time-between-scans-ns TIME_BETWEEN_SCANS_NS
Determines the amount of time it takes for the needle
to return to the starting point to start the next
frame.
--output-non-raster-gif, --no-output-non-raster-gif
Outputs a gif if '--output-non-raster-gif', doesn't if
--no-output-non-raster-gif
--output-raster-gif, --no-output-raster-gif
Outputs a gif if '--output-raster-gif', doesn't if
--no-output-raster-gif
--output-pickle, --no-output-pickle
Outputs a pickle if '--output-pickle', doesn't if
--no-output-pickle
--output-post, --no-output-post
Outputs the post analysis visualizations
--output-path-prefix OUTPUT_PATH_PREFIX
Path to output gif file.
--output-gif-color-map OUTPUT_GIF_COLOR_MAP
Determines the matplotlib color map used for the
output gif. Some useful ones are: 'gist_gray' for
black and white, 'RdBu' for diverging Red-white-blue,
'gist_rainbow' for rainbow.
--output-resolution-x OUTPUT_RESOLUTION_X
x axis Resolution of output gif in pixels. (Up-scaled
from original height Maps size)
--output-resolution-y OUTPUT_RESOLUTION_Y
y axis Resolution of output gif in pixels. (Up-scaled
from original height Maps size)
--output-hdf5, --no-output-hdf5
Outputs a hdf5 file if '--output-hdf5', doesn't if
--no-output-hdf5
--output-hdf5-path OUTPUT_HDF5_PATH
Path to output hdf5 file.
--progress-bar, --no-progress-bar
If true, shows a progress bar.
--enlarge-floaters, --no-enlarge-floaters
If true, spreads floater density based on the radius
in the type name (in angstrom)
--n-cores N_CORES To be used for parallel operation. If not specified,
uses the maximum number of cores.
Using the default arugments, the run time for the demo should be around 10 seconds. The output is a pickle file
containing a dictionary with
3 entries: non-rasterized-maps, a list of numpy arrays containing the height maps for each timepoint,
rasterized-maps, a list of numpy arrays containing the height maps for each timepoint, post rastering,
and args, a dictionary containing the arguments used for the simulation.
A script file, LOCAL_npctransport_sequential.sh is supplied to run the npctransport simulation on the HUJI cluster,
with the required output format. To run the script, use the following command:
csh scripts/LOCAL_npctransport_sequential.sh <from-start(1/0)> <start> <step> <output_folder_path> <config_path> <output_statistics_interval>
Then supply output_path as the input_path argument for the HS-AFM simulation.