This is a readme file for supplemental data for "Computer-generated isotope model achieves experiment-level accuracy of fidelity for position-specific isotope analysis". This includes all the chemical mechanisms generated as well as the code used to simulate the chemistry and create the figures in the paper. Here is a list of the folders and what each contains.
This folder contains the input file, propane_rmg_input_file.py, used by the Reaction Mechanism Generator software (RMG) to generate the full model. With RMG installed, the full isotope model can be run with the command python $rmg/scripts/isotopes.py --kineticIsotopeEffect simple propane_rmg_input_file.py and a mechanism without isotopes can be created with python $rmg/rmg.py propane_rmg_input_file.py. $rmg represents the path to the directory containing RMG-Py source code. You can find more information about how to install and run RMG at rmg.mit.edu.
The propane and methane mechanisms which were generated by RMG and are used in this work are under the folder mechanisms. There are four folders inside this folder called three_reaction_model, six_reaction_model, drg_model, and full_model which corresponds to the four models used in the paper. The file ending indicates the type of file. inp are CHEMKIN input files, cti are Cantera input files, txt are species dictionaries corresponding to the CHEMKIN input file for loading the CHEMKIN input file into RMG, and csv is a datatable comparing the cantera names to properties of isotopologue. All of these files are provided so RMG installation is not necessary to perform analysis.
If the file name contains no_isotopes, then the file contains an RMG run without adding isotopes. If the file name contains KIE, then the file contains isotope labels with kinetic isotope effects. no_KIE indicates isotopes without kinetic isotope effects.
The folder code contains methods for converting between isotopologues and enrichment values and example scripts to help you through the process.
analysis_methods.py contains methods for converting between isotopologues and enrichment values.
cantera_tools.py contains methods for running the cantera module. Comes from the github repository github.com/goldmanm/tools.
prepare_model.py to create a file which maps isotopologues in cantera to specific enrichments, which is necessary for isotopic analysis. Running this file requires RMG to be installed. The output form this file is given in the mechanism sections.
create_paper_figures.py is a script that simulates the models produced by RMG and creates all the figures in the main body of the paper. Code from this file provides an example of how simulations can be run and how to convert from enrichments to isotopologue fractions and vise versa. You must be in the code directory to run this script. Images are saved in the folder results.
The code in this section was tested with Python 2.7 and Python 3.6 with the following packages and their version numbers:
- numpy 1.11.3 (1.15.3 for Py3)
- pandas 1.0.2 (0.23.6 for Py3)
- cantera 2.3.0a3 (2.4.0 for Py3)
- seaborn 0.8 (0.9.0 for Py3)
- statsmodels 0.8.0 (0.9.0 for Py3)
- matplotlib 2.0.2 (2.2.3 for Py3)
If you are not familiar with python, I recommend using anaconda to manage the packages. While most packages are available on anaconda's default channel, the package cantera is not, but it is available on the cantera channel and can be installed with the command conda install -c cantera cantera.
Documentation for running RMG to obtain the mechanisms is available at reactionmechanismgenerator.github.io/RMG-Py/, which is needed for prepare_models.py to run. All other analysis is doable without RMG installation.
This directory lists CSV files containing data from the figures and tables in "Gilbert et al. Measurement of position-specific 13C isotopic composition of propane at the nanomole level. Geochemica et Cosmochemica Acta 177:205-216, 2016". The data originating from figures was approximated using Engauge Digitizer. These are used in creating figures in the paper.
When you run create_paper_figures.py, it will create a results directory and save images here.