Home
.
A high-throughput genome-scale metabolic reconstruction and growth simulation pipeline.
Genome-scale metabolic reconstructions (metabolic 'models') are a representation of all known metabolic reactions that are predicted to occur within an individual cell, plus their associated proteins, genes and metabolites. These models are built via a 'bottom up' approach, leveraging the biomedical literature and public databases to infer the enzymatic profile of a cell based on its genome sequence. I.e. where gene X is known to encode protein Y, which is in turn known to catalyse reaction Z. Once constructed, metabolic models provide a powerful tool to interrogate the cellular biochemistry and predict biological capabilities. See this great review for more details.
Bactabolize is designed to facilitate high-throughput model construction for population comparative analyses. It uses a reference-based approach to generate draft strain-specific models. Blast+ is used to identify gene orthologs in the input genome assembly(ies) as compared to the reference model sequences. By default, orthologs are assigned as the best bi-directional blast hit (BBH) above a pre-determined identity and coverage threshold. The metabolic reactions associated with these orthologs, as well as information about their corresponding proteins and metabolites, are transposed from the reference into the draft model.
Once the models are constructed, Bactabolize can be used to predict their associated growth phenotypes via Flux Balance Analysis (FBA); current applications include growth substrate usage and single-gene deletion analyses.
Bactabolize currently has three main functional modules (draft_model, fba, sgk) as well as a trouble-shooting module (patch_model). The fba and sgk modules can also be applied to models generated via third-party software.
If you use Bactabolize, please cite:
- Vezina B. / Watts S.C. et al. 'Bactabolize: A tool for high-throughput generation of bacterial strain-specific metabolic models'. https://doi.org/10.1101/2023.02.26.530115
- Ebrahim, A., Lerman, J.A., Palsson, B.O. et al. 'COBRApy: COnstraints-Based Reconstruction and Analysis for Python'. BMC Syst Biol 7, 74 (2013). https://doi.org/10.1186/1752-0509-7-74
Bactabolize was developed by Stephen Watts, Ben Vezina and Kelly Wyres. We thank Jane Hawkey, Helena Cooper, Kathryn Holt, Jon Monk and Sylvain Brisse for their important contributions.
We welcome feedback and suggestions. Please use the github issue tracker.