nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

JMP for Immunology?

Phil_Kay

Hi,

We are working with a cell therapy company developing T-cell immunotherapeutic products and vaccines. There is a clear value for DOE in process, formulation and assay development.

Our contact there also thinks there could be value for the Immunology group that they are forming. They will be generating a lot of data from clinical trials and don't have any analysis tools beyond Excel.

They are hoping that we can show a case study demo that will be relevant to the Immunology people. In describing the kind of data that they will have they used terms including: "pharmacokinetic", "antibody responses", "clinical assay readouts" and "measuring action not effect." I think I know what these words mean but I don't know what I would show to prove the value of JMP. I got the sense that they have clinical trial responses that they are tracking over time.

Does anybody know of a relevant demo data set?

Alternatively, let me know if this is not a place for JMP so I don't waste my time.

I've posted part of a job posting, below, for an Immunology Scientist for this company in case that helps.

Thanks,

Phil

Scientist, Immunology

The team is responsible for conducting pre-clinical and clinical
immunology research. We measure antibody and T cell responses to
candidate immunotherapeutic products for the prevention or treatment of
infectious diseases and cancer. The Immunology Scientist will develop,
implement and interpret immunology assays and oversee laboratory operation
and administration. Specific duties include:
• Ensure that antibody and cellular assays are established and appropriately
qualified to support both human and animal studies of our candidates or competing
product candidates
• Ensure all required Technical SOPs are written and in place and maintained for
assays conducted
• Ensure that supervised staff are trained, and training records are up to date.
• Oversee experiments to measure immune responses from animal experiments
and human clinical trials using ELISpot, flow cytometry, ELISA, and other assays
performed internally or externally such as RNA-seq, TCR sequencing, avidity
measurements, surface plasmon resonance, immunohistochemistry
• Ensure clinical assays are conducted according to GCP principles
• Support the tech transfer of required assays to contractors, such as those
conducting toxicology experiments

Accepted answers

karencopeland

I use JMP frequently with diagnostics (i.e., assays of some sort) and clinical trials. What is needed within JMP "depends" on the trial and what the endpoint of that trial is. The trials I have been a part of tend to be smaller as compared to Pharma trials and data collected doesn't necessarily lend itself to JMP Clinical (think 100s not 1000s of subjects with pretty basic data collected).

Items to highlight that JMP is good at include:

1. data import (the clinical data will come from somewhere, often excel but more moving more often to some sort of online EDC (electronic data capture) system so export could be sas transport, cvs, xml, etc.

2. data exploration (recode, distributions at the top of the data file, etc.) to find the messed up clinical data (don't use those words - how about clinical data that may need further monitoring).

3. VIRTUAL JOIN (and regular join). Often clinical data comes in multiple data files, such as a demographic file, a treatment file, lab results file, as well as the assay data that might also be in a separate file.

4. all the basics that they may be looking for.

All comments

Chris_Kirchberg

Hi @Phil_Kay ,

This one is tough from a data point of view. You can use the Cytometry data set for Flow Cytometry analysis. I have a few .fcs files that JMP can import as well. I have not finished a workflow yet to get all .fcs files in, joined, manipulated and then ready for analysis yet.

Bioassay can be used for non-linear/Fit Curve. There is a pharmacokinetic models fit curve option if that helps.

Snee's DOE Clinical Assay example assay should be relevant for DOE and analysis.

Much of this will be fit curve/non-linear modeling.

Dose-Response type analysis would be similar as wel.

@Byron_JMP will have some additional ideas.

Chris_Kirchberg

And one more thought on this. Since data is coming from a clinical trial....what are the medical reviewers using Maybe JMP Clinical can help with that if they do not have it yet.

benfrancis

Hi Phil,

I can only really speak regarding Pharmacokinetics but whilst JMP can fit equations in this area, it doesn't handle (to the best of my knowledge):

patients with different dosing schedules
patients with different sampling schedules
nonlinear mixed effects regression modelling, which is often require to discern between fixed and random effects.

If you hear any of those three, it likely best to recommend in combo with SAS.

https://community.jmp.com/t5/JMP-Scripts/Nonlinear-Degradation-Mixed-Model/ta-p/21367 Not PK but an example of getting SAS PROCNLMIXED and JMP together.

Hope that helps.

Kind Regards,

Ben

ian_jmp

FWIW, I used the Discovery Paper search tool with 'assay' and got eight hits. Worth looking, if you haven't already.

Jeff_Perkinson

You're right, @ian_jmp, there are lots of hits.

2021-02-15_09-12-42.876.png

MikeD_Anderson

Here's a fun paper I've been meaning to work into a demo. It's on cut-point analysis in immunogenicity assays. It's pretty nice in that it has a bunch of different methods shown (all of which appeared to be possible in JMP). More importantly, it includes the data for the examples in the PDF... which also let's you demo PDF import.

1-s2.0-S0022175912003699-main.pdf

Phil_Kay

Thanks @Chris_Kirchberg , @Byron_JMP , @ian_jmp , @benfrancis, @Jeff_Perkinson and @MikeD_Anderson .

This is all really helpful.

I need to check back with the customer. We have plenty of examples on developing and monitoring assays, which I would expect. But I think what they really want is a example that shows how they could look at the clinical data when they run the assays in their trials. I guess it is not a surprise that we don't have that kind of data.

Thanks again,

Phil

Phil_Kay

I think "Cholesterol Stacked" might be useful example.

karencopeland

Items to highlight that JMP is good at include:

2. data exploration (recode, distributions at the top of the data file, etc.) to find the messed up clinical data (don't use those words - how about clinical data that may need further monitoring).

4. all the basics that they may be looking for.

Phil_Kay

Thanks, @karencopeland .

That is really useful. It reminds me that I have worked a lot with scientists working on trials for animal nutrition, which are also smaller trials with some continuous measures along with different end points. And these are exactly the kind of tools that are most useful. Thanks again,

Phil