Bioequivalence and Bioavailability Forum

Main page Policy/Terms of Use Abbreviations Latest Posts

 Log in |  Register |  Search

Back to the forum  2018-07-20 11:02 CEST (UTC+2h)
yjlee168
Senior
Homepage
Kaohsiung, Taiwan,
2008-05-08 05:57

Posting: # 1831
Views: 6,050
 

 stab for R [R for BE/BA]

Dear All,

   Again, we like to announce and share another tool for the stability data analysis using R.
It is called stab for R.
We follow the ICH guideline 'Q1E Evaluation for Stability Data' to design this tool. This guideline describes when extrapolation should be considered as proposing a retest period for a drug substance or a shelf-life of a drug product that extends beyond the period covered by available data from the stability study under the long-term storage condition.
[image]
Thanks.

All the best,
--- Yung-jin Lee & Hsin-ya Lee (KMU, Taiwan)
Helmut
Hero
avatar
Homepage
Vienna, Austria,
2008-05-08 13:36

@ yjlee168
Posting: # 1836
Views: 5,190
 

 stab for R

Dear Yung-jin Lee & Hsin-ya,

again a fine piece of work!

Cheers,
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. ☼
Science Quotes
kegoosse
Junior

Belgium,
2017-08-08 15:01

@ yjlee168
Posting: # 17683
Views: 2,737
 

 stab for R

Indeed a nice tool.

I do have 1 question. My csv file is deemed as an incorrect dataset.
It contains 3 columns with batch, time and assay information (including column titles!)
This, however, doesn't work. How should the csv input file look like?

(PS I already tried to work without column titles. Alas.)
yjlee168
Senior
Homepage
Kaohsiung, Taiwan,
2017-08-08 19:19
(edited by yjlee168 on 2017-08-08 20:27)

@ kegoosse
Posting: # 17684
Views: 2,716
 

 A single-batch drug stability study?

Dear kegoosse,

It happens only if your dataset was obtained from a single-batch drug stability study (or you selected batch type as single-batch from GUI). If this is the case, there is no need to include "batch" in your dataset. Just delete the "batch" column from your dataset and it should work. There only requires two columns of "time" and "assay" for the dataset of a single-batch drug stability. Sorry about this error. I should put more clear information about this in the window of error pop-up dialogue for next release. Finally, the column title is required. Thanks for bringing this error up here.

[edited] if there is still an error, please try to use period(.) as the decimal point instead of comma (,). See this thread for detailed information.
» ...
» (PS I already tried to work without column titles. Alas.)

All the best,
---Yung-jin Lee
[image]bear v2.8.3-1:- created by Hsin-ya Lee & Yung-jin Lee
Kaohsiung, Taiwan http://pkpd.kmu.edu.tw/bear
Download link (updated) -> here
mahmoud-teaima
Regular

2017-08-16 01:18

@ yjlee168
Posting: # 17688
Views: 2,612
 

 stab for R

Dear Yung-jin Lee & Hsin-ya Lee,
Chapeau for the stab package in its new GUI look, a real addition to your group of apps through R.
We all are waiting your update for bear2.8.3 to include the missing pieces of the puzzle (2x3x3 designs, HVD) and (NTI Drugs).
Good luck in all the coming.
Greetings.

Mahmoud Teaima, PhD.
yjlee168
Senior
Homepage
Kaohsiung, Taiwan,
2017-08-16 19:11
(edited by yjlee168 on 2017-08-16 21:54)

@ mahmoud-teaima
Posting: # 17696
Views: 2,610
 

 bear is still going...

Dear Dr. Teaima,

I am working on bear right now with partial replicate BE. So I have already started studying the package replicateBE (v1.0.3 and still not available from CRAN yet) developed by Helmut Schütz, Michael Tomashevskiy, and Detlew Labes. Regarding HVD and NTI, I still don't any plan/clue yet. Any suggestion/comment is welcomed. Once I have finished it, I will first announce the update here. So please stay tuned and thanks for reminding me.

» We all are waiting your update for bear2.8.3 to include the missing pieces of the puzzle (2x3x3 designs, HVD) and (NTI Drugs).

All the best,
---Yung-jin Lee
[image]bear v2.8.3-1:- created by Hsin-ya Lee & Yung-jin Lee
Kaohsiung, Taiwan http://pkpd.kmu.edu.tw/bear
Download link (updated) -> here
Helmut
Hero
avatar
Homepage
Vienna, Austria,
2017-08-16 19:30

@ yjlee168
Posting: # 17697
Views: 2,613
 

 package replicateBE

Hi Yung-jin,

» I have already started studying the package replicateBE v1.0.3 …

Please clone or download the current version (1.05.9000) from GitHub.

» … and still not available from CRAN yet

We haven’t decided yet whether we will submit it to CRAN. ;-)

» Regarding HVD and NTI, I still don't any plan/clue yet.

I reckon a complete implementation of the FDA’s methods is not possible in R.

Cheers,
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. ☼
Science Quotes
ElMaestro
Hero

Denmark,
2017-08-16 20:52

@ Helmut
Posting: # 17698
Views: 2,603
 

 package replicateBE

Hi Hötzi,

» I reckon a complete implementation of the FDA’s methods is not possible in R.

Satterthwaite's denominator df? :-)

if (3) 4

Best regards,
ElMaestro

"(...) targeted cancer therapies will benefit fewer than 2 percent of the cancer patients they’re aimed at. That reality is often lost on consumers, who are being fed a steady diet of winning anecdotes about miracle cures." New York Times (ed.), June 9, 2018.
mittyri
Senior

Russia,
2017-08-16 21:40
(edited by mittyri on 2017-08-16 22:49)

@ ElMaestro
Posting: # 17699
Views: 2,612
 

 FDA RSABE in R is still a Riemann hypothesis

Hi ElMaestro,

» Satterthwaite's denominator df? :-)
Satterthwaite could be solved using lmerTest package.

AFAIK most difficult thing is
REPEATED/GRP=TRT SUB=SUBJ;

I do not give up (yet) with FDA code

PS: will send you some mail

Kind regards,
Mittyri
yjlee168
Senior
Homepage
Kaohsiung, Taiwan,
2017-08-16 21:53

@ Helmut
Posting: # 17700
Views: 2,585
 

 package replicateBE

Dear Helmut,

Thank you so much for your quick response. Now I can see more detailed about the package - replicateBE from GIT. That really helps. I was considering about the output format generated by replicateBE. Will you try to generate SAS-like (or similar at least) outputs in the future (no hurry at all)? I knew that you have done cross-validation with some other software, including SAS. So it seems possible. Let me know if I am wrong.

» ...
» I reckon a complete implementation of the FDA’s methods is not possible in R.

OK. got it.

All the best,
---Yung-jin Lee
[image]bear v2.8.3-1:- created by Hsin-ya Lee & Yung-jin Lee
Kaohsiung, Taiwan http://pkpd.kmu.edu.tw/bear
Download link (updated) -> here
Helmut
Hero
avatar
Homepage
Vienna, Austria,
2017-08-16 23:49

@ yjlee168
Posting: # 17701
Views: 2,595
 

 package replicateBE

Hi Yung-jin,

» I was considering about the output format generated by replicateBE. Will you try to generate SAS-like (or similar at least) outputs in the future (no hurry at all)?

The current output is a hobby-project. However, I added goodies inspectors love and commercial software don’t provide out of the box: Which data was used, who performed the analysis, when, on which system, which OS, which version (of the OS, R, and packages used).

» I knew that you have done cross-validation with some other software, including SAS. So it seems possible.

The main purpose was cross-validation. For that we need only the CVwR and the 100(1–2α) CI. What I use in the function calls is foo(..., details=TRUE). I included more human-readable stuff because beta-testers asked for it. I don’t want to send the model tables to the output. However, if one wants them: foo(..., verbose=TRUE).
Couldn’t resist looking at the Type I Error. Try method.A(data=rds03, ola=TRUE, adj=TRUE) and have a look at the end of the results-file.

Cheers,
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. ☼
Science Quotes
Back to the forum Activity
 Thread view
Bioequivalence and Bioavailability Forum |  Admin contact
18,547 posts in 3,941 threads, 1,192 registered users;
online 24 (1 registered, 23 guests [including 13 identified bots]).

If there is an exception to any rule,
and if it can be proved by observation,
that rule is wrong.    Richard Feynman

The BIOEQUIVALENCE / BIOAVAILABILITY FORUM is hosted by
BEBAC Ing. Helmut Schütz
HTML5 RSS Feed