170 likes | 258 Views
This class project from Tufts University in Fall 2011 focuses on developing approximate expressions for rate changes in metabolic networks based on measurement data under various conditions. It involves problem formulation, solution quality, results analysis, project contributions summary, and code quality assessment.
E N D
White Box Function Estimationusing Convenience Kinetics COMP 150GA – Class Project Fall 2011 Tufts University YOUR NAME, Your.Name@tufts.edu
Goal • Develop approximate expressions for rate change of concentrations in metabolic networks, when given measurement data (data sets that correspond to various operating conditions). • Grading is as follows: • 25% - problem formulation • 25% - solution quality • 30% - results on each test case • 10% - summary of project contributions • 10% - Overall code quality/clarify/documentation/organization • Add as many slides as needed to get your points across
25% Problem Formulation • Explain HOW you formulated the problem as an optimization problem, and JUSTIFY your choices. Please make sure to include an intuitive explanation and a mathematical equation if appropriate. • Define your design space, and explain how you represent a solution in the design space • What is your objective function? • What is your evaluation function? How does it capture trading off accuracy for form simplicity? How are solutions compared? (see next slide)
Evaluation Function • Show and explain how two possible solutions (from any of the examples) were evaluated
25% Your Solution • One slide that provides Outline/Overview/Summary/FlowGraphof Heuristic • Additional slides with details on each key idea
RESULTS • The results should show the following for EACH test case: • One or more plot that shows the accuracy of your best solution against the given data. • One or more plot (or explanation supported by tables/data) that shows how you performed tradeoffs for a particular example. For example, if you did multiple objective optimization, then show or explain a graph of your pareto front, making sure that your axis are clearly explained.
Test Case #1 • D-Glucose + 2 ATP ↔ 2 ADP + FDP ATP ADP FDP
10% Results for Test Case #1
Test Case #2 • S ↔ P
10% Results for Test Case #2
GLUC6P GAP Test Case #3 • GLUC + 2NAD + NADP + ADP ↔ 2Ethanol + 2NADH + NADPH + ATP • GLUC6P + 2NAD + NADP + 2ADP ↔ 2Ethanol + 2NADH + NADPH + 2ATP • GAP + NAD + 2ADP ↔ 2Ethanol + NADH + 2ATP
10% Results for Test Case #3
10% Summary of Project Contributions • What was your project able to accomplish? • If someone else was going to use some of your ideas to solve this problem, which ones would you recommend?
5% BONUS points What Worked and Did not Work • List any clever or seemingly obvious ideas that you tried and did not work.
10% BONUS points Make a skype or in-person appointment with Prof Dec 20 or 21 to review your project