So I got this far already, but how do you show Statistical significance between your IC50 values. I've come to the point where I did a Least squares regression with a best-fit value comparison on IC50 and a sum-of-squares F test to assess statistical significant differences between the different groups it only shows me that the zero hypothesis (no difference) can be rejected and that there is a stat. sign. difference amongst the 4 groups. However I want to know exactly which groups differ from eachother. Could you show how to do a suitable post-hoc analysis to figure this out in graphpad? thanks for considering.
Sorry for taking so long to reply. In GraphPad Prism, IC50 determination involves fitting a dose-response curve to your experimental data to determine the concentration of a compound that inhibits a particular biological process by 50%. The number of parameters used in the IC50 determination depends on the model you choose to fit your data. Three-parameter model: This model assumes a sigmoidal dose-response curve with three parameters: the bottom plateau (representing the response at infinite inhibitor concentration), the top plateau (representing the response at zero inhibitor concentration), and the IC50 value (the concentration at which the response is halfway between the top and bottom plateaus). This model is simpler and more commonly used when you have limited data points or when the curve doesn't exhibit significant asymmetry. Four-parameter model: This model includes an additional parameter, the slope factor (Hill slope), which allows for asymmetry in the dose-response curve. The slope factor represents the steepness of the curve around the IC50 value. This model is more flexible and suitable for data that show significant asymmetry or curvature. When to use each model: Three-parameter model: Use this model when your data follow a standard sigmoidal dose-response curve without significant asymmetry or curvature. This model is simpler and requires fewer data points for reliable fitting. It's a good choice for initial analysis or when the additional complexity of the four-parameter model isn't justified by the data. Four-parameter model: Choose this model when your data exhibit significant asymmetry, curvature, or when you need a more flexible model to accurately capture the dose-response relationship. The four-parameter model provides a more accurate estimation of the IC50 value when the dose-response curve deviates from a simple sigmoidal shape. In summary, start with the three-parameter model for simplicity and use the four-parameter model when your data require a more flexible and accurate fitting approach.
@@bioinfotips yes I used log10. Its a combination of siRNA and drug. So in siRNA lowers the Ic50 of the drug. I also saw this result in apoptosis assay. Also calculating cell viability percent of MTT assay shows that.
So I got this far already, but how do you show Statistical significance between your IC50 values. I've come to the point where I did a Least squares regression with a best-fit value comparison on IC50 and a sum-of-squares F test to assess statistical significant differences between the different groups it only shows me that the zero hypothesis (no difference) can be rejected and that there is a stat. sign. difference amongst the 4 groups. However I want to know exactly which groups differ from eachother. Could you show how to do a suitable post-hoc analysis to figure this out in graphpad? thanks for considering.
Hi Meagan and thank for your comment. Let me search a little bit so i can provide the most accurate answer possible. May also make a video about it.
Also, can you send a print screen of the statistical analysis for the channel email? bioinfotips@gmail.com
Hello. Thank you for the video. How could I make a graph for the IC50 using graphpad prism?
When you do the analysis it automatically plots a graph for you in the graphs section. Check it and let me know if you found it.
Please tell me, which protocol did you use to investigate IC50 ?
Hi 😁 what do you mean?
use a viability assay, like MTS, MTT, WST1 etc
What's the difference between 3 parameters and four parameters? And when should u use either of it?
Sorry for taking so long to reply.
In GraphPad Prism, IC50 determination involves fitting a dose-response curve to your experimental data to determine the concentration of a compound that inhibits a particular biological process by 50%. The number of parameters used in the IC50 determination depends on the model you choose to fit your data.
Three-parameter model: This model assumes a sigmoidal dose-response curve with three parameters: the bottom plateau (representing the response at infinite inhibitor concentration), the top plateau (representing the response at zero inhibitor concentration), and the IC50 value (the concentration at which the response is halfway between the top and bottom plateaus). This model is simpler and more commonly used when you have limited data points or when the curve doesn't exhibit significant asymmetry.
Four-parameter model: This model includes an additional parameter, the slope factor (Hill slope), which allows for asymmetry in the dose-response curve. The slope factor represents the steepness of the curve around the IC50 value. This model is more flexible and suitable for data that show significant asymmetry or curvature.
When to use each model:
Three-parameter model: Use this model when your data follow a standard sigmoidal dose-response curve without significant asymmetry or curvature. This model is simpler and requires fewer data points for reliable fitting. It's a good choice for initial analysis or when the additional complexity of the four-parameter model isn't justified by the data.
Four-parameter model: Choose this model when your data exhibit significant asymmetry, curvature, or when you need a more flexible model to accurately capture the dose-response relationship. The four-parameter model provides a more accurate estimation of the IC50 value when the dose-response curve deviates from a simple sigmoidal shape.
In summary, start with the three-parameter model for simplicity and use the four-parameter model when your data require a more flexible and accurate fitting approach.
Hello. I have a question. My drug A ic50 should be lower than my Drug B Ic50. But in prism it shows opposite. What should I do?
Thank you for your question. Did you use log10 of the concentrations?
Also, in what other software it showed a different result?
@@bioinfotips yes I used log10. Its a combination of siRNA and drug. So in siRNA lowers the Ic50 of the drug. I also saw this result in apoptosis assay. Also calculating cell viability percent of MTT assay shows that.
Are you using the cell viability of MTT as the effect of the drug in the IC50 graph?
Haiii do you know how to count ic50 of two drug that non constant??? Please anybody enlight me if you have the solutionssss... Thxx
Hi :) what do you mean by non constant?
The ratio of two drugs is non constan