How can I see the relation between the parameters-dependent variable and pIC50-independent variablee

annata

Hi everyone!

I'm working on a project about designing a model (QSAR - model) using Deep learning and in the "Deep learning" report i dont see the relation between the parameters and PIC50. Please help me

Thank you.

Garima_Singh

Hi @annata,

Kindly let us know which product is being used for creating the Deep Learning model.

The detail is required to direct the query to the required support team.

Many thanks.

Kind Regards

Garima Singh

annata

Hi Garima_Singh

Thank you for relplying my post, I'm using rapidminer 9.10 for creating Deep learning model

Garima_Singh

Hi @annata,

Thank you for the response.

Hi @RolandJones, kindly support the user on the Rapidminer query.

Kind Regards

Garima Singh

RolandJones

Hi @annata

Thank you for your query, it sounds like an interesting use case! Just to confirm, you're predicting the PIC50 values using a Deep Learning model, and you'd like to extract more details on the relationship between the input variables and the predicted values?

Best,

Roland

annata

Hi RolandJones

Yes, that's exactly what I meant.

Many thanks

Thuy an

RolandJones

Hi @annata,

Given you are predicting a numerical value, I think the Interpretation extension might be a good option here. For example, you can use the "Generate Interpretation" operator to assess the global weights of the model, as well as the absolute local weights.

Please let me know if you need any help implementing this.

Best,

Roland

annata

Hi @RolandJones

When installing "Interpretation" extension, I get this error. Do you know how to solve it?

Thanks,

Thuy An

RolandJones

Could you please share the rapidminer-studio.log file to help diagnose the issue? On Windows, it can be found in %USERPROFILE%\.RapidMiner\rapidminer-studio.log

Thanks,

Roland

annata

Hi @RolandJones

Is it the file you need ?

rapidminer-studio.log

Many thanks,

Thuy An

RolandJones

Hi Thuy an,

Thank you, yes this is what I needed. My apologies, it appears that v0.8.0 of the extension is incompatible with your version of RapidMiner (now AI) Studio - I will raise this with development. You will need to install version v0.7.1 of the extension manually. To do this, please go to https://marketplace.rapidminer.com/UpdateServer/faces/download.xhtml?productId=rmx_interpretation&platform=ANY&version=0.7.1. to download the .jar file. You will then need to place this is in the %USERPROFILE%\.RapidMiner\extensions folder and restart the program. The extension should now work.

Best,

Roland

annata

Hi @RolandJones

It's very kind of you to help me.

Best,

Thuy An

RolandJones

Hi @annata

You're most welcome, if there's anything else that comes up I would be glad to help.

Best,

Roland

annata

Hi @RolandJones

There is one more problem I hope you can show me how to deal with. Well, as you know, I'm creating a deeplearning model and I wonder that how can I save the Deeplearning model?.

To make it more clear, the purpose of my study is about predicting bioactivity of a compound. So, I need a model to predict bioactivity of a data that including compounds.

Like you know, in DL we'll get different results on different runs and I just want to save the best model giving the best result.

I don't know how to save the model, just know how to save the process that generate the model.

Best,

Thuy An

RolandJones

Hi Thuy An,

You can add a "Store" operator and connect the model to store the model in the repository for use later.

Best,

Roland

annata

Hi @RolandJones

You've helped so much. I'm so gratefull <3

Thanks,

Thuy An