53902_en-US

liliconcom

Add-in for determining dose-reduction

When we introduce a new product in biomass, we often characterize this in terms of dose reduction compared to a reference product. I.e. instead of determining that the same amount of enzyme protein is more active by a certain amount, we would like to determine how much less enzyme we can use and still get the same activity.

 

Most often we have done this by a simple reading of the horisontal difference between the two dose-response curves, but this approach does not allow any systematic evaluation of the uncertainty. We have now developed a methodology for this purpuse, which is statistically a bit more sound that the visual reading, and created an add-in that can be used to quickly carry out such an analysis and plot the results. Briefly, the analysis is based on fitting the Hill equation to dose-response data for two or more enzymes.

 

Any inputs for further development are more than welcome. Thanks to MTOF and LCK for their assistance.

Re: Add-in for determining dose-reductionI definitely think we'll chat multiple times next week, for sure!Re: Add-in for determining dose-reduction

@JPKO, I will be happy to give you a little demo next week! You probably need to load the Solver add-in in Excel.

 

Jesper

Re: Add-in for determining dose-reduction

@jfri0: Funny, I was just waiting for someone to tell me they had done this before - after all, dose reductions are a relatively well established basis for comparing enzyme products.

 

I couldn't get your macro to work (something with a library I may not have) or read the documentation (probably something with LUNA), but it could be fun to compare the two approaches. I have a poster on this in the bioenergy poster session during the jamboree next week. It would be great if you can find the time to stop by for a chat. And that goes for you too @Michael_MXAK

Re: Add-in for determining dose-reduction

@JPKO Thanks! I've added the script to NovoJMP Experimental under Analysis: http://gitlab.nzcorp.net/nznarddatascience/NovoJMPAddin/blob/Experimental/NovoJMP.jmpaddin

 

@jfri0 Ew .xls

Re: Add-in for determining dose-reduction

http://nz/fast/link.pl?luna=2002-29536

 

Jesper

Re: Add-in for determining dose-reduction

Glad you liked it! I wouldn't mind at all, if you don't mind my non-optimised sporadically documented code? Also, I have messed a bit around with a couple of alternatives to using the Hill model, and since I posted the add-in I changed it up to use the "mechanical growth" instead because it has one parameter less and apparently fits the data as well. This might be something users would want to change, but I didn't implement it as an option at this point.

 

The full script is pasted below. If you would rather have me deposit it somewere, let me know.

 

Johan

 

// this script will fit a Mecanistic Growth model to any number of dose-response
// data sets with a minimum of 4 observations each, and calculate the percentage 
// difference between the dose needed to achieve a given response. 
//
// JPKO@novozymes.com - 19/10 2017

//
Names default to here(1);
dt=current data table();
dt << clear Select;


// delete previous prediction columns
col_name_list = dt << get column names(string);
if (contains(col_name_list, "Prediction formula"),
 dt<< delete column("Prediction formula" ) 
);

// ------- create dialog to define columns ----------
dlg = Column Dialog(
 Title( "Specify data columns" ),
 groupCol = Col List( "Enzyme", Min Col( 1 ), Max Col( 1 ) ),
 xCol = Col List( "Dose", Min Col( 1 ), Max Col( 1 ), DataType( "Numeric" ) ),
 yCol = Col List( "Response", Min Col( 1 ), Max Col( 1 ), DataType( "Numeric" ) )
);

// check if user decides to quit.
If( dlg["Button"] == -1,
 Throw( "User cancelled" )
);

// process information returned from dialog.
Remove From( dlg );
Eval List( dlg );


dt:groupCol << set property( "Row Order Levels", (1) );


// ----- create dialog to determine the response at which the inverse comparison should be made ------
New Window( "Select response",
 <<MODAL></MODAL>

Re: Add-in for determining dose-reduction

This is nice, and I think would be helpful in 1G Biofuels too! Would you mind if I integrated it into the NovoJMP add-in suite? (http://gitlab.nzcorp.net/nznarddatascience/NovoJMPAddin)

 

If you are okay with it, do you happen to have the original script file available?

 

~Michael

Re: Add-in for determining dose-reduction

Here's an example data table to show the structure needed. Below is a screendump of the generated plot.