You can explore the effect of parameters or infer model predictions by investigating each model you create. Typically, you would look for factor trends, whether they impact the model significantly, or whether they have no effect on your response.
You can read more background details here.
How to investigate the effect of factors on the model
After having created and saved the model of interest, navigate to the Explore Model tab.
From the dropdown menu under Model, on the left hand side of the page, choose the model of interest.
If you want to visualise and investigate the 3D surface of your model response, you can use the 3D plot.
Note: you can select the factors that you want to plot on the x and y axis using the drop down menus beneath the surface profiler.
For the selected model, you will find a Profiler plot for each factor included in the model terms. The Y-axis corresponds to the model’s predicted response and the X-axis corresponds to the factor. The continuous line corresponds to the response behaviour with respect to that factor (profile). The dashed vertical line shows the level of the factor, which can be changed by dragging the slider underneath each plot. Changing a factor level, via the slider, will update the profile of the remaining factors as per the models prediction.
The model’s value of response (displayed in between the Maximize / Minimize buttons and the factor plots) will always update to reflect the predicted response corresponding to the factor values set via the sliders.
Using these plot controls, you can explore the predicted effect of different factor levels on the response and/or on other factors (factor interactions). In the example below, a middle level of IPTG leads to a symmetrical parabola for Glucose. However, a high level of IPTG leads to a parabola skewed to the right for Glucose.
How to optimize a model
When exploring a model, you can choose to search for the maxima or minima of the predicted response. For example, you may be interested in finding factor settings that will lead to highest protein expression (maximum response) or perhaps lowest amount of impurities in your purified product (minimum response). You can do this by clicking either Maximize or Minimize buttons (towards the top of the page, under Profiler). The following items on the page will update accordingly:
The value of response (displayed under the Maximize / Minimize buttons) will show the optimized value.
The sliders corresponding to each factor plot will indicate the factor values at which the optimal response is obtained.
Well done for making it to the end of this tutorial.
To learn how to predict the best conditions from multiple models in Synthace, click here.
To learn how you can use your model to decide on what your next experiments might be, click here.