Guiding the Refinement of Biochemical Knowledgebases with Ensembles of Metabolic Networks and Machine Learning

Medlock, Gregory L.; Papin, Jason A.

doi:10.1016/j.cels.2019.11.006

Cited by 56 publications

(46 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To build these semi-curated reconstructions, we transformed the manually-curated reconstruction using genetic orthology ( Figure 2A , step 3 ) and added all transformed reactions to the recipient de novo reconstruction ( Figure 2A , step 4 ). Lastly, all draft and semi-curated reconstructions were gapfilled using parsimonious flux balance analysis (pFBA)-based gapfilling (Biggs and Papin 2017; Medlock and Papin 2018) to complete biochemical requirements identified in the experimental literature ( Figure 2A , step 5 ) and to produce biomass (see the Supplemental Methods ). As a result, when compared to manually-curated parasite reconstructions (Carey, Papin, and Guler 2017; Abdel-Haleem et al 2018; Chiappino-Pepe et al 2017; Tymoshenko et al 2015), semi-curated reconstructions are larger in scope than de novo reconstructions and generate predictions with comparable accuracy ( Figure 2C-D ).…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Comparative analyses of parasites with a comprehensive database of genome-scale metabolic models

Carey

Medlock

Stolarczyk

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

1 Protozoan parasites cause diverse diseases with large global impacts. Research on the pathogenesis and biology 2 of these organisms is limited by economic and experimental constraints. Accordingly, studies of one parasite are 3 frequently extrapolated to infer knowledge about another parasite, across and within genera. Model in vitro or in 4 vivo systems are frequently used to enhance experimental manipulability, but these systems generally use species 5 related to, yet distinct from, the clinically relevant causal pathogen. Characterization of functional differences 6 among parasite species is confined to post hoc or single target studies, limiting the utility of this extrapolation 7 approach. To address this challenge and to accelerate parasitology research broadly, we present a functional 8 comparative analysis of 192 genomes, representing every high-quality, publicly-available protozoan parasite 9 genome including Plasmodium, 10 and other species. We generated an automated metabolic network reconstruction pipeline optimized for 11 eukaryotic organisms. These metabolic network reconstructions serve as biochemical knowledgebases for each 12 parasite, enabling qualitative and quantitative comparisons of metabolic behavior across parasites. We identified 13 putative differences in gene essentiality and pathway utilization to facilitate the comparison of experimental 14 findings. This knowledgebase represents the largest collection of genome-scale metabolic models for both 15 pathogens and eukaryotes; with this resource, we can predict species-specific functions, contextualize 16 experimental results, and optimize selection of experimental systems for fastidious species. 42hindering drug development, such as resistance to genetic modification. For example, Plasmodium falciparum 43 (malaria) was considered refractory to genetic modification until recently (Ghorbal et al. 2014; Lee and Fidock 44 2014). Entamoeba histolytica (diarrheal disease) has also been refractory to efficient genetic manipulation, and the genomes of Leishmania develop significant aneuploidy under selective pressure (Downing et al. 2011; Sterkers Although these challenges may be circumvented with new technology, the use of clinical samples, and reductionist 1 approaches, little data exist relative to that which is available for most bacterial pathogens. Without adequate 2 profiling data (genome-wide essentiality, growth profiling in diverse environmental conditions, etc.), we do not 3 have the knowledge to rationally identify novel drug targets. Untargeted and unbiased screens of chemical 4 compounds for antiparasitic effects have proven useful (if the parasite can be cultured, e.g. (Jumani et al. 2018; 5 Chao et al. 2018; Love et al. 2017; Subramanian et al. 2018; Lucantoni et al. 2013)), but this approach provides 6 little information about mechanism of action or mechanisms of resistance development. Typical approaches to 7 study drug resistance, such as evolving resistance to identify mutations in a drug's putative target, are not poss...

show abstract

Section: Resultsmentioning

confidence: 99%

“…To gapfill for individual metabolites or biomass (next section), we used a parsimonious flux balance analysis (pFBA)-based approach as originally used in Biggs and Papin (2017) and futher developed in Medlock and Papin (2018). Code is linked in the Supplemental Information.…”

Section: Methodsmentioning

confidence: 99%

Comparative analyses of parasites with a comprehensive database of genome-scale metabolic models

Carey

Medlock

Stolarczyk

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…A) Ensemble FBA performed on glucose or mannose minimal media using an ensemble of 1000 GEMs for Staphylococcus aureus. This ensemble was generated in [6] by iteratively gapfilling a draft reconstruction to enable biomass production in single-carbon source growth conditions supported with experimental data. Mean for the distribution for either condition shown by vertical line of same color.…”

Section: Coupling Ensemble Modeling With Machine Learningmentioning

confidence: 99%

“…Medusa implements a previously-developed algorithm for gapfilling GENREs using growth phenotyping data [6,19]. The algorithm takes a GENRE with an objective function and a dataset of binary growth/no-growth calls on defined media conditions as input.…”

Section: Generating Ensembles From Phenotypic Datamentioning

confidence: 99%

“…This ap- proach is likely to be highly beneficial for studies of organisms for which little biochemical data are available, which typically have many gaps in their GENRE and thus have many highly-variable alternative gap-filling solutions. Although a nascent approach for studying GENREs, we have built on these observations, and ensemble generation and analysis have been applied in several cases [6,19,21,22].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Medusa: software to build and analyze ensembles of genome-scale metabolic network reconstructions

Medlock

Papin

2019

Preprint

Self Cite

View full text Add to dashboard Cite

AbstractUncertainty in the structure and parameters of networks is ubiquitous across computational biology. In constraint-based reconstruction and analysis of metabolic networks, this uncertainty is present both during the reconstruction of networks and in simulations performed with them. Here, we present Medusa, a Python package for the generation and analysis of ensembles of genome-scale metabolic network reconstructions. Medusa builds on the COBRApy package for constraint-based reconstruction and analysis by compressing a set of models into a compact ensemble object, providing functions for the generation of ensembles using experimental data, and extending constraint-based analyses to ensemble scale. We demonstrate how Medusa can be used to generate ensembles, perform ensemble simulations, and how machine learning can be used in conjunction with Medusa to guide the curation of genome-scale metabolic network reconstructions. Medusa is available under the permissive MIT license from the Python Packaging Index (https://pypi.org/) and from github (https://github.com/gregmedlock/Medusa/), and comprehensive documentation is available at https://medusa.readthedocs.io/en/latest/.

show abstract

Synergisms of machine learning and constraint‐based modeling of metabolism for analysis and optimization of fermentation parameters

Khaleghi

Savizi

Lewis

et al. 2021

Biotechnology Journal

View full text Add to dashboard Cite

Recent noteworthy advances in developing high‐performing microbial and mammalian strains have enabled the sustainable production of bio‐economically valuable substances such as bio‐compounds, biofuels, and biopharmaceuticals. However, to obtain an industrially viable mass‐production scheme, much time and effort are required. The robust and rational design of fermentation processes requires analysis and optimization of different extracellular conditions and medium components, which have a massive effect on growth and productivity. In this regard, knowledge‐ and data‐driven modeling methods have received much attention. Constraint‐based modeling (CBM) is a knowledge‐driven mathematical approach that has been widely used in fermentation analysis and optimization due to its ability to predict the cellular phenotype from genotype through high‐throughput means. On the other hand, machine learning (ML) is a data‐driven statistical method that identifies the data patterns within sophisticated biological systems and processes, where there is inadequate knowledge to represent underlying mechanisms. Furthermore, ML models are becoming a viable complement to constraint‐based models in a reciprocal manner when one is used as a pre‐step of another. As a result, a more predictable model is produced. This review highlights the applications of CBM and ML independently and the combination of these two approaches for analyzing and optimizing fermentation parameters.

show abstract

Guiding the Refinement of Biochemical Knowledgebases with Ensembles of Metabolic Networks and Machine Learning

Cited by 56 publications

References 43 publications

Comparative analyses of parasites with a comprehensive database of genome-scale metabolic models

Comparative analyses of parasites with a comprehensive database of genome-scale metabolic models

Medusa: software to build and analyze ensembles of genome-scale metabolic network reconstructions

Synergisms of machine learning and constraint‐based modeling of metabolism for analysis and optimization of fermentation parameters

Contact Info

Product

Resources

About