Analyzing the Potency and Stability of Bioassays

See how to use JMP to analyze biological assays during pharmaceutical and biotechnology development to help prepare for regulatory submissions, develop standard operating procedures and maintain good manufacturing practices. Build and analyze parallelism between a pair of doses and responses, estimate relative potency, determine when a drug dose or concentration reaches toxicity, determine the affinity of the enzyme to the substrate, identify the 50% dose concentration (EC50) necessary to cause half of the maximum possible effect, and determine equivalence of different formulations.

Questions answered by Valerie Nedbal and Chris Kirchberg at the live webinar:

Q: How do you suggest putting in biological triplicate values for each dose in JMP?

A: You can enter them as rows for each dose.

Q: Is the inflection point analogous to EC50?

A: It is relative to EC50, to be more specific. But absolute will be 50% of max (upper asymptote).

Q: Please explain the  Difference between F-Test parallelism and Chi Square test Parallelism.

A: The F test provides a p-value that is superior to 0.05 while the chi-square test does the opposite. See parallelism test details.

Q: Is there a way to do a pairwise comparison for each curve against the standard alone?

A: Yes, if you want to limit to standard only for relative potency.

Q: In the presentation, Parallelism is assessed by the F test.  For us, the equivalence test- ratios of asymptotes and B - is used more often to assess parallelism . Does JMP offer this approach or any other Parallelism test?

Q: Can we compare JMP to PLA ((Parallel-Line-Assay) as a potency assay analysis tool?

A: You can use compare parameter estimates option to compare growth rate for the linear portion of the curve, if that is what you are looking for.

Q: Is there any option to get the 95% CI values for Relative potency

A: JMP currently only gives the standard error.

Q: Why is 25% difference considered acceptable for Equivalence Test?

A: Here is why we chose 25% , but you can change it.

Q: Can JMP calculate % RP from multiple plates?

A: Each test could be from a different plate all put into one data table.

Q: Is data put in  JMP manually or can it be moved there from the acquisition software electronically in the part 11 compliant manner?

A: Yes, if the data source is ODBC-compliant, JMP can connect to databases and pull data using ODBC drivers or HTTPS Request () and it would be in memory and not in a file to do the analysis.

Q: Is JMP GxP compliant, meaning there are different levels at which different people can execute calculations and data management?

A: Here is a link to our quality statement. Validation is left to the organization using it. JMP is an exploratory tool and SOPs drive much of what GxP strives for. The link provided allows for IQ and OQ validation to assist in GxP environments.

Q: Can you do ANOVA?

A: It depends on how you want to do that . You can use Fit Y by X for simple ANOVA. But if the ANOVA is for stability analysis, or degradation, there is a whole separate platform.  See the video below.

Q: When doing parallelism test, you used 25% range. Apparently, that can also be other ranges, such as 10%.  How will you choose that range to judge if a parameter is different between a sample and the standard?

A: Did you mean Equivalence Test? Parallelism Test does not use a range. For Equivalence Tests, there is a red triangle next to each one that lets you choose those lines.

Resources

• Nedbal blog, How to perform Bioassay analysis with JMP
• Statistical Details for Fit Curve Models
• Microtiter Well Plate Map Shape Definition File that includes  examples of the most common well plates, 6 well, 12, 24, 48, and 96, in addition to some larger plates with 384, 1536 and 9600 wells plus two files JMP needs to recognize the well names and build graph builder charts like the one above.