ramdiv/ber-model

Computational model of Base-Excision-Repair (BER) mechanism

This repository contains the implementation of the computational model of BER mechanism, as described in [paper]

Requirements

• KaSim 3.5
• Python 2.7
• simplejson

Usage

Main entry point is the "run.py" file. It requires the user to specify the run mode as it's first command line argument, e.g:

$ python run.py [mode]

where mode is one of the following:

• synth, which synthesizes sample DNA strands. It supports optional named argument -descriptor that points to the JSON settings file.
• simulate, which simulates the BER mechanism. It supports optional named argument -descriptor that points to the JSON settings file.
• experiment, which runs a series of experiments. It requires an additional argument that points to a JSON file that describes the experiment.

All values defined in the JSON files can be overridden in command line.

synth and simulate mode have a set of predefined default values. The supplied JSON file (-descriptor) needs not to define every value, instead, missing values are filled from the default dictionary.

JSON files are separated into "meta" and "data" top-level keys. "meta" defines the "meta" values, used to direct output, etc. "data" is instead used to define values used directly in the simulation.

synth

• meta:temp_dir

  defines the temporary directory used in the synthesis

• meta:out_dir

  defines the output directory

• meta:out

  specifies the filename of the synthesized DNA

• data:bp_per_strand

  defines the number of base pairs per strand

• data:strands

  defines the number of strands in the output

• data:weights

  defines the dictionary of weights used when generating the DNA

simulate

• meta:temp_dir

  defines the temporary directory used in the simulation

• meta:out_dir

  defines the output directory

• meta:out

  specifies the filename of the trace of the simulation

• meta:dna_file

  specifies the filename of the DNA file, on which the BER is performed

• meta:time / meta:events

  specifies the amount of "simulation time"/events used for the simulation

• meta:plot_points

  specifies the number of points in the trace of the simulation

• meta:model

  is an array that defines the components used in the simulation:

  • damage -- damages the DNA file after 1000 events
  • sliding -- sliding action for the DG actor
  • DG, APE1, POLb, LIG3, XRCC1, PNKP -- actions for the respective actor
  • xrcc_dimer XRCC1 dimerization action

• meta:sanity

  used for debugging. Outputs weakly compressed stories of some observables.

• data:...

  specifies the kinetic rates used in the simulation -- see rates.xlsx

experiment

meta fields are joined with the ones defined before (in simulate).

• meta:experiment_name

  defines the name of the experiment series

• meta:runs

  specifies the number of times the rates are calculated and simulations are performed

• meta:parallel

  specifies the number of simulations to perform in parallel

• data:constants

  specifies the dictionary of constants used in calculation of rates

• data:fields

  specifies the dictionary of fields to generate the rates for. The keys of the dictionary should match the names of the rate variables. The value of each key should be another dictionary with two components type and value.

  Currently supported types are: constant, random_uniform, Kd_constraint.

  • constant copies the value of the value key as the value of the rate variable
  • random_uniform generates a uniform random value in the range specified in the value key. The range is defined as a JSON array of three values, the first one is either "int" or "float", followed by the lower bound, and the upper bound. If the first value is missing, "float" is used. You can also use the name of a variable defined in the constants section, if the variable defines an array of the form specified before.
  • Kd_constraint, used to impose the K_d constraint on variables that are generated. The value field should be an array, with the rate variable name used as k_off in the first position, followed by Kd ratio.