Towards the network-based prediction of repurposed drugs using patient-specific metabolic models

Pacheco, Maria Pires; Bintener, Tamara Jean Rita; Sauter, Thomas

doi:10.1016/j.ebiom.2019.04.017

Cited by 8 publications

(6 citation statements)

References 10 publications

(11 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This relationship has yet to be fully understood, and therefore predicting cellular growth following genetic manipulations is still challenging. Understanding and controlling cellular growth have important applications in disease modeling, biotechnology, and for the development of efficient cell factories (13). CRISPR-Cas9-enabled genetic engineering now gives the ability to modify yeast DNA with single-nucleotide precision in vivo (14), achieving engineered strains that maximize a desired output.…”

Section: Significance Statementmentioning

confidence: 99%

A mechanism-aware and multiomic machine-learning pipeline characterizes yeast cell growth

Culley

Vijayakumar

Zampieri

et al. 2020

Proc. Natl. Acad. Sci. U.S.A.

View full text Add to dashboard Cite

Metabolic modeling and machine learning are key components in the emerging next generation of systems and synthetic biology tools, targeting the genotype–phenotype–environment relationship. Rather than being used in isolation, it is becoming clear that their value is maximized when they are combined. However, the potential of integrating these two frameworks for omic data augmentation and integration is largely unexplored. We propose, rigorously assess, and compare machine-learning–based data integration techniques, combining gene expression profiles with computationally generated metabolic flux data to predict yeast cell growth. To this end, we create strain-specific metabolic models for 1,143 Saccharomyces cerevisiae mutants and we test 27 machine-learning methods, incorporating state-of-the-art feature selection and multiview learning approaches. We propose a multiview neural network using fluxomic and transcriptomic data, showing that the former increases the predictive accuracy of the latter and reveals functional patterns that are not directly deducible from gene expression alone. We test the proposed neural network on a further 86 strains generated in a different experiment, therefore verifying its robustness to an additional independent dataset. Finally, we show that introducing mechanistic flux features improves the predictions also for knockout strains whose genes were not modeled in the metabolic reconstruction. Our results thus demonstrate that fusing experimental cues with in silico models, based on known biochemistry, can contribute with disjoint information toward biologically informed and interpretable machine learning. Overall, this study provides tools for understanding and manipulating complex phenotypes, increasing both the prediction accuracy and the extent of discernible mechanistic biological insights.

show abstract

Section: Significance Statementmentioning

confidence: 99%

A mechanism-aware and multiomic machine-learning pipeline characterizes yeast cell growth

Culley

Vijayakumar

Zampieri

et al. 2020

Proc. Natl. Acad. Sci. U.S.A.

View full text Add to dashboard Cite

show abstract

“…The potential of such models for the investigation of optimal processing is now acknowledged and practiced 10 , 37 – 40 . Examples for applicative use include the optimal production or utilization of industrial compounds such as xylose 41 , biofuels 42 , vitamins 43 and drug development 44 . In bioremediation, genome scale metabolic modeling approaches were applied for the design of Geobacter sulfurreducens strain capable of increased electron transfer and a higher Fe(III) respiratory that was beneficial for environmental bioremediation of uranium-contaminated groundwater 45 .…”

Section: Introductionmentioning

confidence: 99%

Genome-scale reconstruction of Paenarthrobacter aurescens TC1 metabolic model towards the study of atrazine bioremediation

Ofaim

Zarecki

Porob

et al. 2020

Sci Rep

View full text Add to dashboard Cite

Atrazine is an herbicide and a pollutant of great environmental concern that is naturally biodegraded by microbial communities. Paenarthrobacter aurescens TC1 is one of the most studied degraders of this herbicide. Here, we developed a genome scale metabolic model for P. aurescens TC1, iRZ1179, to study the atrazine degradation process at organism level. Constraint based flux balance analysis and time dependent simulations were used to explore the organism's phenotypic landscape. Simulations aimed at designing media optimized for supporting growth and enhancing degradation, by passing the need in strain design via genetic modifications. Growth and degradation simulations were carried with more than 100 compounds consumed by P. aurescens TC1. In vitro validation confirmed the predicted classification of different compounds as efficient, moderate or poor stimulators of growth. Simulations successfully captured previous reports on the use of glucose and phosphate as biostimulators of atrazine degradation, supported by in vitro validation. Model predictions can go beyond supplementing the medium with a single compound and can predict the growth outcomes for higher complexity combinations. Hence, the analysis demonstrates that the exhaustive power of the genome scale metabolic reconstruction allows capturing complexities that are beyond common biochemical expertise and knowledge and further support the importance of computational platforms for the educated design of complex media. The model presented here can potentially serve as a predictive tool towards achieving optimal biodegradation efficiencies and for the development of ecologically friendly solutions for pollutant degradation. Atrazine (2-chloro-4-ethylamino-6-isopropylamino-1,3,5-triazine) is an herbicide employed to control broadleaf and grass mainly in crops such as rice, wheat, maize, and sorghum. It is also a well-known pollutant of great environmental concern. Atrazine has been shown to have negative effects such as DNA damage, gene expression shifts, cancer and endocrine disruption 1,2. Its residues are found in soil samples decades after it was last applied and were shown to chronically leach into local aquifers 3. As such, efforts are being made to limit and monitor its use 4. While atrazine was banned in the European Union and Switzerland since 2003, the United States Environmental Protection Agency still allows its wide use under monitoring 5,6. Areas contaminated with atrazine and other hazardous herbicides are rapidly increasing worldwide introducing a need in remediation approaches. Bioremediation-an environmental bioprocess in which naturally occurring organisms are used for breaking down hazardous substances into less toxic or non-toxic substances-is increasingly acknowledged as a cost-effective feasible alternative for environmental cleaning 7-9. Critical environmental bioprocesses are naturally carried out by bacteria and are related to removal of pollutants from water, soil or air 10. Thus, bacterial bioremediation is widely applied for the ...

show abstract

“…This task is of practical relevance in many biomedical contexts and might pave the way for the development of automated strategies for experimental hypothesis generation. In particular, the introduction of our framework contributes to the emerging field of approaches combining sample-specific metabolic modeling with machine learning to classify cancer samples and/or to predict drug response, as recently reviewed [ 49 , 50 ].…”

Section: Discussionmentioning

confidence: 99%

On the Use of Topological Features of Metabolic Networks for the Classification of Cancer Samples

Machicao

Craighero

Maspero

et al. 2021

View full text Add to dashboard Cite

Background: The increasing availability of omics data collected from patients affected by severe pathologies, such as cancer, is fostering the development of data science methods for their analysis. Introduction: The combination of data integration and machine learning approaches can provide new powerful instruments to tackle the complexity of cancer development and deliver effective diagnostic and prognostic strategies. Methods: We explore the possibility of exploiting the topological properties of sample-specific metabolic networks as features in a supervised classification task. Such networks are obtained by projecting transcriptomic data from RNA-seq experiments on genome-wide metabolic models to define weighted networks modeling the overall metabolic activity of a given sample. Results: We show the classification results on a labeled breast cancer dataset from the TCGA database, including 210 samples (cancer vs. normal). In particular, we investigate how the performance is affected by a threshold-based pruning of the networks by comparing Artificial Neural Networks, Support Vector Machines and Random Forests. Interestingly, the best classification performance is achieved within a small threshold range for all methods, suggesting that it might represent an effective choice to recover useful information while filtering out noise from data. Overall, the best accuracy is achieved with SVMs, which exhibit performances similar to those obtained when gene expression profiles are used as features. Conclusion: These findings demonstrate that the topological properties of sample-specific metabolic networks are effective in classifying cancer and normal samples, suggesting that useful information can be extracted from a relatively limited number of features.

show abstract

Towards the network-based prediction of repurposed drugs using patient-specific metabolic models

Cited by 8 publications

References 10 publications

A mechanism-aware and multiomic machine-learning pipeline characterizes yeast cell growth

A mechanism-aware and multiomic machine-learning pipeline characterizes yeast cell growth

Genome-scale reconstruction of Paenarthrobacter aurescens TC1 metabolic model towards the study of atrazine bioremediation

On the Use of Topological Features of Metabolic Networks for the Classification of Cancer Samples

Contact Info

Product

Resources

About