1. From genomes to Pathway tools and from Pathway Tools to Metabolic Models Jeremy Zucker Broad Institute of MIT and Harvard

2. Outline Enzyme prediction with EFICAz TBCyc curation FungiCyc curation Network debugging tools

3. Pathway predictor Enzyme predictor Model Predictions Omics data From genome sequences to Metabolic flux models

4. Enzyme Prediction EFICAz: – Predicts Enzyme Commision Numbers – Functionally Discriminating Residues Homofunctional MSA Heterofunctional MSA – Support Vector Machine – Sequence Identity Threshold – Integrates 9 independent methods to achieve maximal accuracy KEGG-BLAST

5. Enzyme Prediction EficazTool incorporated into Calhoun web interface Runs on LSF cluster for all Transcripts in an AnalysisRun Automatically assigns EC number OntologyTerm and ToolEvidence to Transcript

6. TBCyc curation

7. TBDB  TBCyc integration Palsson Flux models => curated pathways McFadden flux model => curated pathways Literature => curated pathways Expression data => Omics viewer Chip-Seq => curated gene regulatory network Coming soon: – annotation jamboree for TB – Connect to Decode Open source drug discovery for TB

8. NeurosporaCyc curation

9. Neurospora curation Resources: – Model organism – Active Community Annotation Project – Recent genome assembly and annotation – Systems Biology grant Chip-seq KO library Biofuels

10. Biofuels from Neurospora? Growing interest for obtaining biofuels from fungi Neurospora crassa has more cellulytic enzymes than Trichoderma reesei N. crassa can degrade cellulose and hemicellulose to ethanol [Rao83] Simultaneous saccharification and fermentation means that N. crassa is a possible candidate for consolidated bioprocessing Xylose Ethanol

11. Effects of Oxygen limitation on Xylose fermentation in Neurospora crassa Zhang, Z., Qu, Y., Zhang, X., Lin, J., March 2008. Effects of oxygen limitation on xylose fermentation, intracellular metabolites, and key enzymes of Neurospora crassa as3.1602. Applied biochemistry and biotechnology 145 (1-3), 39-51. Xylose Pyruvate TCAEthanol RespirationFermentation Glycolysis Oxygen level (mmol/L*g) Ethanol conversion (%) Low O 2 Intermediate O 2 High O 2

12. Pentose phosphate Aerobic respiration Fermentation TCA Cycle Model of Xylose Fermentation Xylose Oxygen Ethanol ATP Two paths from xylose to xylitol

13. Pentose phosphate Aerobic respiration Fermentation TCA Cycle Oxygen=5 ATP=16.3 NADPH Regeneration NADPH & NAD + Utilization High Oxygen NAD + Regeneration

14. Pentose phosphate Aerobic respiration Fermentation TCA Cycle Ethanol Low Oxygen Oxygen=0

15. Pentose phosphate Aerobic respiration Fermentation TCA Cycle Ethanol Intermediate Oxygen Optimal Ethanol NADPH & NAD Utilization Oxygen=0.5 ATP=2.8 NAD Regeneration NADPH Regeneration All O 2 used to regenerate NAD used in first step

16. Pentose phosphate Aerobic respiration Fermentation TCA Cycle Ethanol Intermediate Oxygen Optimal Ethanol NADPH & NAD Utilization Oxygen=0.5 ATP=2.8 NAD Regeneration NADPH Regeneration All O 2 used to regenerate NAD used in first step Bottleneck Pyruvate decarboxylase Improve NADH enzyme

17. Xylose  Lipids col-2 (Glucose-6-phosphate dehydrogenase) mutant produces lipids from xylose and glucose – 35x more TAGS than WT – 12% biomass is lipid Mapping out the metabolic pathways to explain this phenomenon. Lipids for NADPH regeneration?

18. Algorithms for Debugging Metabolic Networks Metabolic network is too complicated. The metabolic network is infeasible. E-flux results in dead model.

19. Minimal Organism Given a feasible model under the given nutrient conditions, find the fewest number of nonzero fluxes that still results in a viable organism. minimize card(v) subject to: Sv = 0, l <= v <= u

20. Minimal Reaction Adjustment Given an infeasible model, find a reaction with the smallest number of reactants and products that results in a feasible model. minimize card( r ) subject to Sv + r = 0 l <= v <= u

21. Minimal Limit Adjustment Given a set of (feasible) baseline limits, and an (infeasible) set of expression-constrained flux limits, find the smallest number of adjustments to the flux limits that results in feasible model (without exceeding the baseline constraints). minimize card(dl) + card(du) subject to: Sv = 0, l – dl <= v <= u+du l-dl >= l_0, u+ du <= u_0 dl >= 0, du >= 0.

22. Minimum Cardinality Each of these problems is a special case of the minimum cardinality problem Minimize card( x ) Subject to Ax + By >= f Cx + Dy >= g Caveat: the minimum cardinality problem is NP-Hard!

23. Sparse Optimization to the rescue! Recent results from Compressed Sensing have shown that minimizing the L1-norm is a decent heuristic Iterative methods can improve the results Instead of minimize card( x ), Minimize norm(diag(w) x, 1) = sum(w_i |x_i| ) Update w_i = 1/(epsilon + |x_i|), i=1,…,k

24. Implementations 3 software packages in matlab – Cvx (Sedumi and spdt3) – Glpkmex (GLPK) – Cplexmex (CPLEX) min cardinality – MILP and L1 heuristic version min limit adjustment – MILP and L1 heuristic version Min limit adjustment => min cardinality

25. Acknowle

26. Acknowledgements Broad Institute James Galagan Bruce Birren Brian Haas Aaron Brandes Matt Henn Li Jun Ma Christina Cuomo Carsten Russ Broad Genome Sequencing Platform Program Project Heather Hood SRI Peter Karp Markus Krumenacker Suzanne Paley Mario Latendresse Tomer Altman… Finishing Team Margaret Priest Harindra Arachchi Lynne Aftuck Mike Fitzgerald Genome Assembly Sarah Young Sean Sykes Annotation Team Brian Haas Mike Koehrsen Qian Zeng Tom Walk

27. Outline e354f1709c ICS Guest network