AW: linear relation
Hello,
thank you for your answers.
My problem is, that I tried to fit an inverse sigmoidal model, but I just have
few concentrations in the array around IC50. Physiologically an EMAX of 1 does
not make sense in my case.
But now I think a linear model wouldn't help me either.
Von: Rik Schoemaker [mailto:[email protected]]
Gesendet: Donnerstag, 14. Oktober 2010 08:27
An: Hauke Rühs
Betreff: RE: [NMusers] linear relation
Dear Haucke,
Are you absolutely sure there is no alternative to linear? Why not stick to the
usual solutions:
INH=1-C1/(EC50+C1)
or
INH=1-(C1**H/(EC50**H+C1**H))
That way you avoid the nasty break at 1 and of course you don't need
conditions...
Good luck!
Rik
_____
Quoted reply history
From: [email protected] [mailto:[email protected]] On
Behalf Of Hauke Rühs
Sent: 13 October 2010 5:30 PM
To: [email protected]
Subject: [NMusers] linear relation
Dear NMusers,
I currently building an indirect response model
C1=A(1)/V1
DADT(2)=KIN-KOUT*INH*A(2)
, where the KOUT is inhibited by INH=1-(SLOPE*C1). SLOPE should be estimated
with variability. How can I make sure that INH stays between 0 and 1?
I wrote it like this
IF ((SLOPE*C1).GT.1) THEN
INH=0
ENDIF
IF ((SLOPE*C1).LT.0) THEN
INH=1
ENDIF
IF((SLOPE*C1).LE.1.AND.(SLOPE*C1).GE.0) THEN
INH=1-SLOPE*C1
ENDIF
, but I think it is not optimal to write it in conditions. Does anyone know a
better way to write this?
Thank you.
-----------------------------
Hauke Rühs
Apotheker
Pharmazeutisches Institut
- Klinische Pharmazie -
An der Immenburg 4
53121 Bonn
Tel: + 49-(0)228 73-5781
Fax: + 49-(0)228 73-9757
http://www.klinische-pharmazie.info/ www.klinische-pharmazie.info