I have a simple optimization to be done statistically using PB. I have to study the effect of medium composition on cell mass. I have to check the effect of a variety of variables, eg C sources like monosachs, disacchs, polysacchs, etc. and nitrogen sources like organic and inorganic. Therefore, I want to study the effect of the above variables independently, is it possible to do that using PB or would RSM be of help?
Yes you can do it independently. Only you need to do the experiments with minimum standard deviation (i.e. repeat atleast 5-6 times) and secondly you need to know the dummies in the analysis very carefully. Since I have no idea about the medium you are using or any other parameters, I can not comment further. By all means you should be able to finish the optimization within a months time. Good luck
Dear Sir,
mine is a simple medium with 1 C and 1 N source.
As per the stuff that I have refered to, I must be using 2 factorial design. however i am unable to design the expt. is the design create by software or I construct it manually by assigning high and low values to my variables. how do I decide onto the number of dummy variebles and the constituents of the same. I would highly appreciate to recieve some refernce material or valuable comments
Recently i have did RSM study.
Which software are you using? Minitab or design expert.
The tutorials available in the software. First do screening of variables by selecting either half or full fractional factorial design. which determines the number of experiments in duplicates. Then give your high and low values. The design is created by software according to your coded values.
Dear Aparna
I think Naveen has really done an excellent job. That attached paper will be very helpful to you and it is referred by many beginneers. Now coming back to your question on number of variables; see the medium composition and decide the components which has least effect on the yield. As you say your medium is 1c and 1N type, which still leaves plenty of room to decide on the dummies.
from variety of carbon and Nitrogen source you can select best contributed carbon and nitrogen source by PBD. After that to choose the concentration of these parameter you should go Response surface modelling of Design expert software.
if you need help then give me parameter and and your screened out result i can give you design.
thank you evryone for your valuable comments. I am now able to visualize the design to a great extent. I will get back with my queries (if any) during the course of run.
Thank you all!!
i think this will help you
https://www.youtube.com/watch?v=PJFXInzvdtQ
Dear all,
I have two questions regarding PB design:
i. How to know which are my dummy variables for my experiment? Are they decided based on the results obtained or set during the experiment is designed?
ii. How to account for standard errors in case of replicate measurements of responses. I found from the literature that we can take the response as the average of the replicates. Then how to cater for the standard error of measurement?
Thanking you
Prasenjit
- Badges
- Science topic
- Similar topics
- Biological Science
- Microbiology