Modelseed 2.0: A Next Generation Resource for Automated Model Reconstruction, Comparison, and Analysis

Genome-scale metabolic models continue to drive our understanding of cellular life, with applications ranging from phenotype prediction[1,2] to drug design[3-5]. In 2010, ModelSEED, an online resource for the reconstruction, exploration, comparison, and analysis of genome-scale metabolic models, was released to address the need for high-throughput metabolic model reconstruction[6].  Over the years, ModelSEED has grown exponentially, currently containing over 60,000 models constructed by over 5,000 users.  Although regarded as a prominent web-based resource for metabolic modeling, ModelSEED 1.0 faces a barrier of entry for biologists, as well as issues of extensibility, performance, and usability.

Here we present ModelSEED 2.0, available for early preview at http://modelseed.theseed.org.  ModelSEED 2.0 is a single-page web application built on emerging, open-source web services, including ProbModelSEED (Probabilistic-ModelSEED).  The reengineering of core services and UI has improved existing functionality, performance, and usability. This application paves a path for the analysis of ‘omics data and the dissemination of knowledge through sharing and experiential education.

The new ModelSEED allows authenticated users to reconstruct models, gapfill, and run FBA simulations on public, private, and shared genomes/models with a variety of options.  On the fly editing of biomass, media, and model reactions, combined with a curated and searchable biochemistry database, makes model editing quick and intuitive with this application.  In addition to comprehensive data views, individual/comparative analysis of users’ or collaborators’ models is easily achieved through the click of a few buttons. SVG and HTML5 canvas heatmaps, visualizations, and pathway maps derived from KEGG[7, 8] provide a familiar, yet modern web experience for the comparison of gene presence, gapfilled reactions, FBA, and more.  Gene/reaction knockouts via metabolic networks enable users to ask “what if” without the distraction of “how”, eliminating what is often a complex task with other web resources. The new ModelSEED also comes preloaded with extensive public data, including over 50 published genome-scale metabolic models, as well as draft models for over 20K public genomes.

 

1. Ohno, S., H. Shimizu, and C. Furusawa, FastPros: screening of reaction knockout strategies for metabolic engineering. Bioinformatics, 2014. 30(7): p. 981-7.
2. Tepper, N. and T. Shlomi, Predicting metabolic engineering knockout strategies for chemical production: accounting for competing pathways. Bioinformatics, 2010. 26(4): p. 536-43.
3. Cornish-Bowden, A. and M.L. Cardenas, Metabolic analysis in drug design. C R Biol, 2003. 326(5): p. 509-15.
4. Tang, J. and T. Aittokallio, Network pharmacology strategies toward multi-target anticancer therapies: from computational models to experimental design principles. Curr Pharm Des, 2014. 20(1): p. 23-36.
5. Schmidt, B.J., J.A. Papin, and C.J. Musante, Mechanistic systems modeling to guide drug discovery and development. Drug Discov Today, 2013. 18(3-4): p. 116-27.
6. Henry, C.S., et al., High-throughput generation, optimization and analysis of genome-scale metabolic models. Nature biotechnology, 2010. 28(9): p. 977-982.
7. Kanehisa, M., et al., Data, information, knowledge and principle: back to metabolism in KEGG. Nucleic Acids Res, 2014. 42(Database issue): p. D199-205.
8. Kanehisa, M. and S. Goto, KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res, 2000. 28(1): p. 27-30.