# Team:Illinois-Tools/Modeling

(Difference between revisions)
 Revision as of 06:16, 21 July 2009 (view source)Pcdoshi2 (Talk | contribs)← Older edit Latest revision as of 03:46, 22 October 2009 (view source)Damfahr2 (Talk | contribs) (→The Algorithm) (14 intermediate revisions not shown) Line 6: Line 6: .firstHeading {display: none;} .firstHeading {display: none;} table { table { - background-color: white; + background-color: #FFFFFF; - font-color: white; + font-color: black; color:white; color:white; } } a.menu { a.menu { - background-color: white; + background-color: #FFFFFF; - color: white; + color: black; width: 12em; width: 12em; } } .firstHeading { .firstHeading { - color:white; + color:black; } } #bodyContent { #bodyContent { - background-color: white; + background-color: #FFFFFF; } } #content { #content { - background-color: #00008B; border: none; + background-color: white; } } Line 32: Line 32: } } p { p { - color:white; + color:black; } } body { body { - background: #00008B url(https://static.igem.org/mediawiki/2009/2/2b/IllinoisToolsBackground.gif) repeat-x; + background: white url(https://static.igem.org/mediawiki/2009/2/2b/IllinoisToolsBackground.gif) repeat-x; } } Line 41: Line 41: + [[Image:Modelingillinoistools.gif]] Line 47: Line 48: - - 889 + + + =='''Modeling'''== + + The Illinois - Tools team can model any pathway, whose starting and ending compounds are stored in the Kegg database.  The algorithm developed by the team takes information from the Kegg database and finds the most optimum pathway, based on the weights selected by the user.  Examples include the pathway with the least number of steps, or the pathway that uses the least amount of ATP. + + The Illinois-Tools team also wishes to use this algorithm to help other IGEM teams in modeling their own desired pathways.  For example, the our program can help the Illinois wetlab team in modeling their pathway.  The wetlab team's project is about a binary decoder in the organism E. coli, that senses 2 inputs, such as 2 sugars, and produces one of four possible outputs, which are fluorescent proteins, based on the combination of inputs. + + Go to [https://2009.igem.org/Team:Illinois Illinois] + + A possible future expansion of their project would be to model these pathways computationally, and see if it would be experimentally feasible. Furthermore, once the protein output is obtained, they want to check if it can be used to produce useful compounds like biofuels. + + To help this team reach its goals, our program, IMPtools, was able to model a pathway from D-Arabinose to Ethanol in E. coli.  The result can be seen below.  It has been optimized for E.coli, and can be further optimized when the wetlab team is ready to work on it.  The labels on the reaction are the Kegg IDs for compounds. + + [[Image:Illinoistoolsarabinosetoethanolpathway.jpg|center|]] + + + ===The Algorithm=== + + In modeling for other teams, a few things should be considered. The algorithm requires interactivity. It is can only optimize if you set the parameters to specify what you mean by "optimal". In that light, the user of IMPtools when modeling for a practical design application should consider how important removing excess reactants or creating biproducts could be, as well as the implications of ATP consumption. Sometimes the algorithm can return intermediate compounds that are truly only cofactors of a main reaction (these can generally be avoided by weighting against high order nodes, but also by specifically removing certain nodes). With IMPtools and a small amount of manipulation, a reasonable pathway can be returned using IMP's pathfinding algorithm. Using these results, actual experimentation is necessary to validate the pathway - perhaps, given future feedback the algorithm could account for user's successes and failures and preferentially bias results toward those experimentally validated.

## Modeling

The Illinois - Tools team can model any pathway, whose starting and ending compounds are stored in the Kegg database. The algorithm developed by the team takes information from the Kegg database and finds the most optimum pathway, based on the weights selected by the user. Examples include the pathway with the least number of steps, or the pathway that uses the least amount of ATP.

The Illinois-Tools team also wishes to use this algorithm to help other IGEM teams in modeling their own desired pathways. For example, the our program can help the Illinois wetlab team in modeling their pathway. The wetlab team's project is about a binary decoder in the organism E. coli, that senses 2 inputs, such as 2 sugars, and produces one of four possible outputs, which are fluorescent proteins, based on the combination of inputs.

Go to Illinois

A possible future expansion of their project would be to model these pathways computationally, and see if it would be experimentally feasible. Furthermore, once the protein output is obtained, they want to check if it can be used to produce useful compounds like biofuels.

To help this team reach its goals, our program, IMPtools, was able to model a pathway from D-Arabinose to Ethanol in E. coli. The result can be seen below. It has been optimized for E.coli, and can be further optimized when the wetlab team is ready to work on it. The labels on the reaction are the Kegg IDs for compounds.

### The Algorithm

In modeling for other teams, a few things should be considered. The algorithm requires interactivity. It is can only optimize if you set the parameters to specify what you mean by "optimal". In that light, the user of IMPtools when modeling for a practical design application should consider how important removing excess reactants or creating biproducts could be, as well as the implications of ATP consumption. Sometimes the algorithm can return intermediate compounds that are truly only cofactors of a main reaction (these can generally be avoided by weighting against high order nodes, but also by specifically removing certain nodes). With IMPtools and a small amount of manipulation, a reasonable pathway can be returned using IMP's pathfinding algorithm. Using these results, actual experimentation is necessary to validate the pathway - perhaps, given future feedback the algorithm could account for user's successes and failures and preferentially bias results toward those experimentally validated.