385091_en-US

liliconcom

High Throughput Screening

What do you like about this case study?

• It is from R&D - most of our science and engineering case studies for EDA and Predictive Modelling are from Manufacturing.
• The data comes from a journal article (attached .pdf) so it is real data.
• You can show very compelling visuals - "wow" factor.
• You can show simple, powerful ways to extract important insight that would not be possible with Excel.
• You can show that JMP Pro methods outperform standard modelling even for fairly small data.

 

 

Who is it for?

• It should be very relevant for anyone in Chemicals or small molecule Pharma development.
• Anyone from R&D should be able to see how it could translate to their situation. 

 

How should I use it?

1. You can use either the 150 row subset or the full data to demo EDA as per the Tablet demo. The attached slide gives a back story from a customer situation that you might like to use.
2. You can use it to demo Predictive Modelling in JMP Pro. The attached journal and the video below show an example of this type of demo. Use Model Screening to build models and then Formula Depot to apply the best model to the full data and see how well it performs. The model is not perfect but it does a remarkably good job of predicting all 4608 possibilities from a subset of just 150!

 

What platforms can we use?

• Distribution
• Graph Builder
• Partition
• Model Screening
• Formula Depot
• More...