gene overexpression...

[choi]

i have a question about gene expression. 

i know it's impossible in COBRA toolbox

nevertheless, i heard that is is possible similarly to realize gene expression in COBRA toolbox.

it is that a paticular gene relate to reaction (e.g 'G6PDH2r') simply regulated as command fuction ' model = changeRxnBounds(model, 'G6PDH2r', -100, 'l')'

purpose of experiment is that 'G6PDH2r' is effective to secretion of 'EX_gdpfuc(e)' as regenerating NAPD(H)

so i try to simulation FBA for secretion of 'EX_gdpfuc(e)' 

model = readCbModel;

model = changeRxnBounds(model, 'EX_glc(e)', -100, 'l');

model = changeRxnBounds(model, 'EX_o2(e)', -18.5, 'l');

model = changeRxnBounds(model, 'G6PDH2r', -50, 'l');

model = changeObjective(model, 'EX_gdpfuc(e)');

FBAsolution = optimizeCbModel(model);

but if i input or not 'G6PDH2r' as command fuction, it is unchanged value.

please give me a advise... 

choi.

[Siu Hung Joshua Chan]

Hi Choi,

Please give more details about your original model. What is the upper/lower bound for G6PDH2r originally? What is the flux for G6PDH2r and EX_gdpfuc(e) respectively? Do they have fluxes or not? 

Do you mean -100 for the lower bound originally and then you changed it to -50? 

(you said 'it is that a paticular gene relate to reaction (e.g 'G6PDH2r') simply regulated as command fuction ' model = changeRxnBounds(model, 'G6PDH2r', -100, 'l')' and then you tried 'model = changeRxnBounds(model, 'G6PDH2r', -50, 'l');') 

If nothing has been changed, normally it should be 1000 for upper bound and -1000 for lower bound. 

In either case, by changing the bound to -50 you are actually constraining the model, i.e. underexpressing the gene rather than overexpressing.

Do you want to overexpress it or underexpress it?

If you do not have any information on the upper/lower bound for the original model, you normally have -1000 and 1000 as lower and upper bounds. In this case, you can't really predict the results of overexpressing a certain gene. Except in the first place you constrain the upper/lower bounds for a reaction in your original model, for example, using some experimentally measured value from enzyme assay (say -10) in your wild type. Or use other ways to estimate. And then you can try to release the bound (say -50) to predict the case for overexpression.

Best,

Joshua

choi於 2015年1月29日星期四 UTC+1上午9時04分02秒寫道：

[Siu Hung Joshua Chan]

And you said the value had no change no matter you changed the bound or not.

Have you checked in your original model whether G6PDH2r has flux > -50 or > the original lower bound ?

In the case of constraining the lower bound (if your original lower bound is more negative than -50, and then you set it to -50), if the flux > -50 originally, then you won't expect changes because the same optimal solution keeps feasible in your new model.

In the case of releasing the lower bound (if your original lower bound is less negative than -50, then you set it to -50), if the flux is again > the original bound, then similarly even you releasing the bound, it won't change because it is already optimal originally without hitting the lower bound. This may be somehow interpreted as overexpressing the gene being not useful to have more products if the bounds in your original model are set to be biologically relevant.

Best,

Joshua

Siu Hung Joshua Chan於 2015年1月29日星期四 UTC+1下午3時20分11秒寫道：

[choi]

thank you for your help :). joshua.

it is my mistake that value of 'G6PDH2r' is two value

it is arbitrary value... i wanted to input value '-100'

i talk again to you more detail...

purpose of experiment is whether or not 'G6PDH2r' and 'GOFUCR' have an effect on secretion of 'EX_gdpfuc(e)' as regenerating NAPD(H)

gdp-fucose pathway is as follows.

D-glu(e) -> PGI -> PFK -> MAN6PI -> PMANM -> MAN1PT2 -> GMAND -> GDMANE -> GOFUCR -> EX_gdpfuc(e)

i simulated FBA as follows.

model = readCbModel;

model = changeRxnBounds(model, 'EX_glc(e)', -100, 'l');

model = changeRxnBounds(model, 'EX_o2(e)', -18.5, 'l');

model = changeObjective(model, 'EX_gdpfuc(e)')

FBAsolution = OptimizeCbModel(model, 'max');

printFluxVector(model, FBAsolution.x, true)

as a result, 'EX_gdpfuc(e)' is 14.9022mmol/gDW

but 'G6PDH2r' is zero value. i think that it is not involved in gdp-fucose pathway

so i simulate FBA with regard to gene overexpression.

'G6PDH2r' -> lower bound = -1000, upper bound = 1000

'GOFUCR' -> lower bound = -1000, upper bound = 1000

model = changeRxnBounds(model, 'G6PDH2r', -1000, 'l');

model = changeRxnBounds(model, 'GOFUCR', -1000, 'l');

FBAsolution = optimizeCbModel(model, 'max')

printFluxVector(model, FBAsolution.x, true);

'EX_gdpfuc(e)' is 14.9022...

'G6PDH2r' is zero value and 'GOFUCR' is 14.9022...

Have any method that gene overexpression can realize in these condition? 

sorry, joshua... i read repeatedly your opinion, but i can't applied them...