Skip to content

Latest commit

 

History

History
180 lines (142 loc) · 5.82 KB

README.md

File metadata and controls

180 lines (142 loc) · 5.82 KB
Raw

Dependencies

Python

Used the following versions.

  • Python 3.6
  • Pandas 0.20.3
  • Matplotlib 2.2.2

Matlab

  • Matlab > 2015

The following scripts should be present in the Matlab path.

  • Legendflex package from https://github.com/kakearney/legendflex-pkg
  • Save2PDF from https://uk.mathworks.com/matlabcentral/fileexchange/16179-save2pdf

Data type explanations

Summary files (summary_*.tsv)

The columns are as follows:

  • CRN the CRN number.
  • unique is the CRN the exemplar for a specification isomorphic class.
  • one the unoptimized score (rate 1.0).
  • score the best optimized rate.
  • r0 reaction 0 optimised rate.
  • r1 reaction 1 optimised rate etc.

Heatmap files (Bimol_*.tsv)

The columns are as follows:

  • Column 0 input s0 initial value.
  • Column 1 input s1 initial value.
  • Column 2 unoptimized score (rate 1.0).
  • Column 3 optimized score.

Hitting time files (Bimol_*.time)

The columns are as follows:

  • i1 initial values of species s1 etc.
  • r0 optimized rate for reaction r0 etc.
  • opt_time absorption time of rates (r0, r1 etc)
  • one_time absorption time with rates all 1.0.

Time-accuracy trade off files (Bimol_*.speedtime)

The columns are as follows:

  • i0 initial values of species s0. same for i1 etc.
  • r0 the optimized rate of reaction 0, same for r1 etc.
  • opt_time the absorption time with optimised rates.
  • one_time the absorption time with rate 1.0.
  • score the accuracy of with the optimized rates.

Figures

Figure 2

Figure showing all the pre and post optimisation scores for AM_{3,3} and AM_{4,4} CRNs. Also shows the top scoring AM_{4,4} CRNs and shows how they compare to the top two AM_{3,3} CRNs.

Data:

  • 3 reactions 3 species Approximate Majority results - AMno11_S3_R3/summary_archetype.tsv

  • 4 reactions 4 species Approximate Majority results - AMno11_S4_R4/summary_archetype.tsv

  • 4 reactions 4 species Approximate Majority results (specification isomorphic only) extra columns marking presence of different sub-networks. - AMno11_S4_R4/summary_unique_subnets.tsv

Code:

  • python paper/fig_am_overview/overview_am_horiz.py

Figure 3

Heat maps of selected AM_{3,3} and AM_{4,4} CRNs.

Data

  • paper/AMno11_S3_R3/Bimol_28.lbs - LBS file with reactions
  • paper/AMno11_S3_R3/Bimol_28.time - Hitting time data
  • paper/AMno11_S3_R3/Bimol_28.tsv - Heat map
  • paper/AMno11_S3_R3/Bimol_36.lbs
  • paper/AMno11_S3_R3/Bimol_36.time
  • paper/AMno11_S3_R3/Bimol_36.tsv
  • paper/AMno11_S4_R4/Bimol_3750.lbs
  • paper/AMno11_S4_R4/Bimol_3750.time
  • paper/AMno11_S4_R4/Bimol_3750.tsv
  • paper/AMno11_S4_R4/Bimol_4854.lbs
  • paper/AMno11_S4_R4/Bimol_4854.time
  • paper/AMno11_S4_R4/Bimol_4854.tsv

Code

Uncomment the appropriate directory in lines 8 to 13.

  • matlab Matlab/create_heatmaps.m

Figure 4

Figure showing the phase space and sepaatrix for two AM_{3,3} CRNs.

  • matlab Matlab/run_separatrix.m

Figure 5

Figure showing how long it takes to find CRNs with different numbers of reactions and species.

Data

This data format is as follows

  • Column 0 is the CRN number
  • Column 1 is $K + 1$ where the CRN was found
  • Column 2 is the time in seconds (since process start) when the CRN was found.

The files used are

  • paper/AMno11_z3_timmings/ApproximateMajorityNo11_3_3.tsv
  • paper/AMno11_z3_timmings/ApproximateMajorityNo11_4_4.tsv
  • paper/AMno11_z3_timmings/ApproximateMajorityNo11_4_3.tsv
  • paper/AMno11_z3_timmings/ApproximateMajorityNo11_3_4.tsv

Code

  • matlab Matlab/z3_am_time.m

Figure 6

Showing how CME calculations scale with species numbers.

Data

The data for this figure is stored in

  • Matlab/CMECalcEfficiency/times1.mat

Code

  • matlab Matlab/CMECalcEfficiency/cme_am_time.m

Figure 7

Maximum 4 species 3 reactions, pre and post optimisation histograms.

Data

  • paper/maximum_out_S4_R3/summary_archetype.tsv

Code

  • python paper/fig_max_speed_ac/fig_max_histo.py

Figure 8

Maximum 4 species 3 reactions, Example speed/accuracy trade off figure.

Data

  • paper/maximum_out_S4_R3/Bimol_1.speedtime - Speed-accuracy trade off data
  • paper/maximum_out_S4_R3/Bimol_1.lbs - CRN file used
  • paper/maximum_out_S4_R3/Bimol_1_tf10000.tsv - heatmap file

Code

  • python paper/fig_max_speed_ac/fig_max_histo.py

Figure 9

Figure showing all the Division CRNs and how they score.

Data

These are the results of optimising all the Division CRNs.

  • paper/DivNsel_S3_R3/summary.tsv
  • paper/DivNsel_S6_R2/summary.tsv
  • paper/DivNsel_S3_R5/summary.tsv
  • paper/DivNsel_S3_R2/summary.tsv
  • paper/DivNsel_S5_R4/summary.tsv
  • paper/DivNsel_S4_R5/summary.tsv
  • paper/DivNsel_S5_R3/summary.tsv
  • paper/DivNsel_S4_R2/summary.tsv
  • paper/DivNsel_S4_R3/summary.tsv
  • paper/DivNsel_S5_R2/summary.tsv
  • paper/DivNsel_S3_R4/summary.tsv
  • paper/DivNsel_S6_R3/summary.tsv
  • paper/DivNsel_S4_R4/summary.tsv

Code

  • python paper/fig_div/fig_div_sumary.py

Figure 10

Selected heatmaps to display division CRN behaviour.

Data

The heat maps and CRNs selected for accuracy heatmap.

  • DivNsel_S4_R4/Bimol_61586.tsv - Heatmap
  • DivNsel_S4_R4/Bimol_61586.lbs - CRN code
  • DivNsel_S4_R4/Bimol_79523.tsv
  • DivNsel_S4_R4/Bimol_79523.lbs
  • DivNsel_S5_R4/Bimol_751168.tsv
  • DivNsel_S5_R4/Bimol_751168.lbs
  • DivNsel_leader/Bimol_2.tsv
  • DivNsel_leader/Bimol_2.lbs

Code

Uncomment the appropriate directory in lines 8 to 13.

  • matlab Matlab/create_heatmaps.m

CRN Storage

All CRNs shown in the paper are stored in a compressed format in the Storage directory. To decompress all CRNs of a particular specification, you can call (for example)

python python/lbs_storage.py --crn_storage_file Storage\AMno11_S3_R3_CRNs.tsv -d AMno11_S3_R3 -r all

Alternatively, you can choose to decompress a single CRN by using its integer identifier as the argument for r.