Analysis of Metabolic Network Disruption in Engineered Microbial Hosts due to Enzyme Promiscuity

• Vladimir Porokhin, Department of Computer Science, Tufts University, Medford, MA, vladimir.porokhin@tufts.edu
• Sara A. Amin, Department of Computer Science, Tufts University, Medford, MA, sara.amin@tufts.edu
• Trevor B. Nicks, Department of Chemical and Biological Engineering, Tufts University, Medford, MA, trevor.nicks@tufts.edu
• Venkatesh Endalur Gopinarayanan, Department of Chemical and Biological Engineering, Tufts University, Medford, MA, venkatesh.endalur_gopinarayanan@tufts.edu
• Nikhil U. Nair, Department of Chemical and Biological Engineering, Tufts University, Medford, MA, nikhil.nair@tufts.edu
• Soha Hassoun, Department of Computer Science and Department of Chemical & Biological Engineering, Tufts University, Medford, MA, soha.hassoun@tufts.edu

Multicopy Suppressors

Multicopy suppressor cases are analyzed using the patrick-fba.py script. It requires the following files to be present in the data directory:

• reactions-patrick.tsv (included): list of promiscuous reactions generated by PROXIMAL
• iML1428-iso_Glucose.json: iML1428 model from Supplementary Dataset 1 in [1]
• kegg-list-eco.txt: list of E. coli genes from KEGG.
• kegg-link-eco-ec.txt: list of E. coli enzymes linked to genes, also from KEGG.

The script should be run twice, using FBA and MOMA separately:

python patrick-fba.py fba
python patrick-fba.py moma

These commands can be executed in parallel. The script generates tables as in the Supplementary File 1 accompanying the paper and writes them to the output directory.

[1]: Monk, J., Lloyd, C., Brunk, E. et al. iML1515, a knowledgebase that computes Escherichia coli traits. Nat Biotechnol 35, 904-908 (2017). https://doi.org/10.1038/nbt.3956

3-HP Synthesis Pathways

Processing the 3-HP synthesis pathways is done using two scripts. The first one, 3hp-fba.py, is responsible for conducting the simulations of the baseline, engineered, and disrupted models following the method described in the paper. The following files must to be present in the data directory to run the script:

• data/reactions-3hp.tsv (included): list of promiscuous reactions generated by PROXIMAL
• data/iML1515.json: iML1515 model from the BiGG website.

The script needs to be run three times for each pathway (p1 and p2): for scenario 1 (s1), scenario 2 (s2), and both combined (fixed). Thus, a full analysis requires running six commands as follows:

python 3hp-fba.py p1 s1 fba %seed%
python 3hp-fba.py p1 s2 fba %seed%
python 3hp-fba.py p1 fixed fba %seed%
python 3hp-fba.py p2 s1 fba %seed%
python 3hp-fba.py p2 s2 fba %seed%
python 3hp-fba.py p2 fixed fba %seed%

These commands can all be executed in parallel to maximize CPU utilization. Results of the simulations will be placed in the output directory.

Once all simulations are completed, 3hp-fig.py, can be used to compile the results into scatter plots similar to the ones presented in the paper:

python 3hp-fig.py

The figures will be placed in the output directory.