Reproducing JMP Genomic Scripts in JMP Pro

[#1]

Hi, I am transitioning from running microarray data in JMP Genomics to JMP Pro. 

I routinely use the predefined JMP Genomics scripts such as:

Analytical Processes/Workflows/Basic/Basic Expression Workflow or

Analytical Processes/Expression/Normalization/Data Standardize procedures.

I was able to reproduce the STD normalization steps of the Normalization script but not the Basic Expression workflow.

One reason being because it is hard to identify all the steps used the analysis in order to calculate the same columns needed to create the volcano plots (example of data table attached from "View Data" output), cluster & heat map and PCA plots (attached screenshot). Another reason is that I find instructions to use "Response Screening" options of JMP Pro under "Analyze/Screening/Response Screening" or "Analyze/Fit Model/Personality: Response Screening" but none of them have all options available. The first gives the volcano plots but do not allow complex models and the second allow complex models (but not random variables) but don't give the volcano plots. Both of them show p-value cut-off of 0.01 and 0.05 for LogWorth but graphs are made only for p=0.01. Is there a way to change it to 0.05 or define another p-value? Is there a way to reproduce the same batch analysis included in the "Basic Expression Workflow" in JMP Pro?

[#2]

Hi @JeandO ,

Welcome to the community.

Some of the options in the basic expression workflow were not ported to JMP Pro. Specifically the normalization steps.  Product Management has acknowledged this request and I added you to the customer list requesting this feature.

The other items were implemented incrementally as versions and people request them. For example, Response Screening within Fit Model allows for more complex models, but not with random effects in version 17. In version 18, random effects will be added.

For now, some of what you seek has to be performed manually. Like taking the Response Screening output (from either Analyze>Screening or Analyze>Fit Model) and generating/saving the tables (means comparison tables) as a separate data table, then use graph builder to make the graphs and add custom reference lines.

Other times like Hierarchical Clustering and PCA (as well the sample QC part with Density as Y and Intensity Bins as X) can be done using Fit Y by X, but it takes a few steps.

Overall, I too really liked the workflows in JMP Genomics and the current method to try and recreate the steps is to use the Workflow Builder to create your own workflow to be re-used later.

Putting some of the items you mentioned on the Wish List will help Product Management prioritize what to add, where and when (also gives others a chance to vote as well).

I started using Projects to contain all my steps of analysis, but I could have used Workflow Builder to capture the scripts created in the background so I could have used them again later. I am not sure if you have seen this webinar, but take a look at an example analysis that I did that I could have captured as a workflow in workflow builder that would mimic a workflow in JMP Genomics.

Some more posts about gene expression analysis in JMP Pro can be found here.

We are continuing to enhance JMP Pro to contain features like you have requested and we thank you for your patience as we add them.