Re: [NMusers] Logic Expression in $DATA: ACCEPT or IGNORE

[Xinting Wang]

Dear Katya and Bill,

Thanks a lot for your suggestion. Multiple IGNORE, as suggested Katya,
actually worked.

Best Regards

On 18 November 2014 17:24, Ekaterina Gibiansky 
wrote:

>  Hi Xinting,
>
> You can separate the condition into several statements. For example, if
> you need to accept (A=1 OR A=2) AND B<100
> you instead can have 2 or 3 IGNORE statements (depending on the values A
> can have), say
>  IGNORE=(B.GE.100)  IGNORE=(A.GT.2) IGNORE=(A.LT.1)
>
> Regards,
> Katya
>
> Ekaterina Gibiansky, Ph.D.
> CEO&CSO, QuantPharm LLC
> Web: www.quantpharm.com
> Email:   egibian...@quantpharm.com
>
>
> On 11/18/2014 3:24 PM, Denney, William S. wrote:
>
>  Hi Xinting,
>
>
>
> I’ve worked with these types of statements a good bit, and my personal
> preference is to add a column to the data set that makes the selection
> simpler (e.g. set it to 1 if (A == 1 or A == 2) and B < 100).  Last I knew,
> it wasn’t possible to do an “AND” in an ignore statement (and checking the
> $data documentation in NONMEM 7.2, that still appears to be the case).
>
>
>
> Thanks,
>
>
>
> Bill
>
>
>
> *From:* owner-nmus...@globomaxnm.com [mailto:owner-nmus...@globomaxnm.com
> ] *On Behalf Of *Xinting Wang
> *Sent:* Tuesday, November 18, 2014 2:10 PM
> *To:* nmusers@globomaxnm.com
> *Subject:* [NMusers] Logic Expression in $DATA: ACCEPT or IGNORE
>
>
>
> Dear all,
>
>
>
> I am having a difficulty in writing a logic expression of accepting a
> complex structure under $DATA session.
>
>
>
> For example:
>
>
>
> ACCEPT=c(A=1,A=2)
>
>
>
> means that record of A equals to 1 or 2 would be accepted. However, for a
> logic expression below:
>
>
>
> (A=1 OR A=2) AND B<100
>
>
>
> which would require A equals to 1 or 2 and in the mean time, B are less
> than 100.
>
>
>
> NONMEM manual VIII did not specify this type of acceptance. Could anybody
> please tell me how to write this kind expression using ACCEPT, or IGNORE in
> nonmem? Thanks a lot for your help.
>
>
>
> Best Regards
>
>
>
> --
>
> Xinting
>
>
>


-- 
Xinting

Re: [NMusers] Genotype data missing in some individuals

[Jeroen Elassaiss-Schaap]

Dear SoJeong,

First you might want to answer the question whether that phenotype is indeed 
important in your dataset. With the initial popPK model you could plot posthoc 
clearance against bodyweight and/or inspect the posthocs of clearance for 
evidence of multiple peaks in your distribution. You also may see the impact of 
phenotype in stratified concentration versus time plots. Depending on the 
dataset, with its sampling scheme, number of subjects (perhaps a low number) 
and distribution across age, it could be masked. 

If the impact is clear however, it might be benificial to try to include the 
subjects wih missing genotype. With a clear effect, you might be able to 
develop a mixture model. The mixture  approach would describe the different 
populations in your dataset corresponding to the different phenotypes. The 
genotype would than inform the mixture as a covariate - the missing information 
would fall back to the pure mixture approach. As a warning, this approach is 
quite difficult. I would advise you to read up on the nonmem guides ($MIX) on 
this and look in the literature for examples - the Karlsson group has published 
about it, most recently this one (it contains code): 
http://link.springer.com/article/10.1208/s12248-009-9093-4. A search in the 
literature gives you additional background such as 
http://www.page-meeting.org/pdf_assets/9595-PAGE2007_3.pdf and 
http://link.springer.com/article/10.1007/s10928-006-9038-9. 

If the impact is not clear, a more empirical approach might be called for, in 
this case a subset analysis, i.e. where you exclude the missing subjects, of 
the covariate relationship might be all that you could achieve. If there is no 
impact at all, you do not need the genotype of course.

Hope this helps!

Best regards,

Jeroen

http://pd-value.com
jer...@pd-value.com
@PD_value
+31 6 23118438
-- More value out of your data!



On Nov 19, 2014, 7:57 AM, at 7:57 AM, "이소정"  wrote:
>Dear all,
>
> 
>
>I’ve analyzed a tacrolimus PopPK in pediatric patients.
>
>As you know, CYP3A5 genotype can change the tacrolimus PK
>significantly, 3A5 genotyping was performed in the study, 
>
>however, in 20% of the subjects, the genotype data was missed. 
>
> 
>
>Then, how can I reflect the CYP3A5 genotype effect to the tacrolimus
>population model appropriately?
>
>Is there any solution?
>
> 
>
>Best regards,
>
>SoJeong Yi
>
> 

RE: [NMusers] Genotype data missing in some individuals

[Rekic, Dinko]

Dear SoJeong,

I agree with everything Jeroen proposed. In addition to that, you may want to 
code the subjects with missing genotype as genotype 99 (or something similar) 
and then estimate genotype as categorical covariate on CL. This approach is not 
elegant but it is quick and often useful for initial analysis.

Kind regards
Dinko

"The contents of this message are mine personally and do not necessarily 
reflect any position of the Government or the Food and Drug Administration."



From: owner-nmus...@globomaxnm.com [mailto:owner-nmus...@globomaxnm.com] On 
Behalf Of Jeroen Elassaiss-Schaap
Sent: Wednesday, November 19, 2014 6:16 AM
To: 이소정
Cc: nmusers@globomaxnm.com
Subject: Re: [NMusers] Genotype data missing in some individuals


Dear SoJeong,

First you might want to answer the question whether that phenotype is indeed 
important in your dataset. With the initial popPK model you could plot posthoc 
clearance against bodyweight and/or inspect the posthocs of clearance for 
evidence of multiple peaks in your distribution. You also may see the impact of 
phenotype in stratified concentration versus time plots. Depending on the 
dataset, with its sampling scheme, number of subjects (perhaps a low number) 
and distribution across age, it could be masked.

If the impact is clear however, it might be benificial to try to include the 
subjects wih missing genotype. With a clear effect, you might be able to 
develop a mixture model. The mixture  approach would describe the different 
populations in your dataset corresponding to the different phenotypes. The 
genotype would than inform the mixture as a covariate - the missing information 
would fall back to the pure mixture approach. As a warning, this approach is 
quite difficult. I would advise you to read up on the nonmem guides ($MIX) on 
this and look in the literature for examples - the Karlsson group has published 
about it, most recently this one (it contains code): 
http://link.springer.com/article/10.1208/s12248-009-9093-4. A search in the 
literature gives you additional background such as 
http://www.page-meeting.org/pdf_assets/9595-PAGE2007_3.pdf and 
http://link.springer.com/article/10.1007/s10928-006-9038-9.

If the impact is not clear, a more empirical approach might be called for, in 
this case a subset analysis, i.e. where you exclude the missing subjects, of 
the covariate relationship might be all that you could achieve. If there is no 
impact at all, you do not need the genotype of course.

Hope this helps!

Best regards,

Jeroen

http://pd-value.com
jer...@pd-value.com
@PD_value
+31 6 23118438
-- More value out of your data!
On Nov 19, 2014, at 7:57 AM, "이소정" 
mailto:sjlph...@gmail.com>> wrote:
Dear all,


I’ve analyzed a tacrolimus PopPK in pediatric patients.
As you know, CYP3A5 genotype can change the tacrolimus PK significantly, 3A5 
genotyping was performed in the study,
however, in 20% of the subjects, the genotype data was missed.


Then, how can I reflect the CYP3A5 genotype effect to the tacrolimus population 
model appropriately?
Is there any solution?


Best regards,
SoJeong Yi

Re: [NMusers] Genotype data missing in some individuals

[Leonid Gibiansky]

I would do mixture model only if there is a very large -several folds- 
difference in PK parameters for two genotypes. If the difference is 
comparable with the inter-subject variability within the genotype, I 
would introduce category "missing" to remove the effect of those 
subjects on covariate effect estimate. So if the genotype is binary 
(YES/NO), you introduce the new third level "missing", work with it as 
with the 3-level categorical covariate, and report the difference 
between NO and YES as the genotype effect on PK. As a check for 
consistency, you may want to check whether the estimate of the PK 
parameter for "missing" level is somewhere between the estimates for 
"NO" and "YES" levels, closer to the value for the level with higher 
prevalence in your dataset.

Regards,
Leonid

--
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web:www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel:(301) 767 5566



On 11/19/2014 6:16 AM, Jeroen Elassaiss-Schaap wrote:

Dear SoJeong,

First you might want to answer the question whether that phenotype is
indeed important in your dataset. With the initial popPK model you could
plot posthoc clearance against bodyweight and/or inspect the posthocs of
clearance for evidence of multiple peaks in your distribution. You also
may see the impact of phenotype in stratified concentration versus time
plots. Depending on the dataset, with its sampling scheme, number of
subjects (perhaps a low number) and distribution across age, it could be
masked.

If the impact is clear however, it might be benificial to try to include
the subjects wih missing genotype. With a clear effect, you might be
able to develop a mixture model. The mixture  approach would describe
the different populations in your dataset corresponding to the different
phenotypes. The genotype would than inform the mixture as a covariate -
the missing information would fall back to the pure mixture approach. As
a warning, this approach is quite difficult. I would advise you to read
up on the nonmem guides ($MIX) on this and look in the literature for
examples - the Karlsson group has published about it, most recently this
one (it contains code):
http://link.springer.com/article/10.1208/s12248-009-9093-4. A search in
the literature gives you additional background such as
http://www.page-meeting.org/pdf_assets/9595-PAGE2007_3.pdf and
http://link.springer.com/article/10.1007/s10928-006-9038-9.

If the impact is not clear, a more empirical approach might be called
for, in this case a subset analysis, i.e. where you exclude the missing
subjects, of the covariate relationship might be all that you could
achieve. If there is no impact at all, you do not need the genotype of
course.

Hope this helps!

Best regards,

Jeroen

http://pd-value.com
jer...@pd-value.com
@PD_value
+31 6 23118438
-- More value out of your data!

On Nov 19, 2014, at 7:57 AM, "이소정" mailto:sjlph...@gmail.com>> wrote:

Dear all,

I’ve analyzed a tacrolimus PopPK in pediatric patients.

As you know, CYP3A5 genotype can change the tacrolimus PK
significantly, 3A5 genotyping was performed in the study,

however, in 20% of the subjects, the genotype data was missed.

Then, how can I reflect the CYP3A5 genotype effect to the tacrolimus
population model appropriately?

Is there any solution?

Best regards,

SoJeong Yi

No virus found in this message.
Checked by AVG - www.avg.com 
Version: 2014.0.4765 / Virus Database: 4189/8594 - Release Date: 11/18/14

RE: [NMusers] Genotype data missing in some individuals

[Denney, William S.]

Hi SoJeong,

I agree with Leonid here on the value of the mixture model.  With potentially 
subtle changes, mixture models can be very difficult.  One way that I've had 
luck previously with a similar approach is to make "unknown genotype" a 
separate category and then to fit a parameter that is fraction "yes" (similar 
to a mixture model, but not specifying a genotype for a subject).  Something 
like:

G1 = THETA(1)
G2 = THETA(2)
FRA = 1/(1+EXP(-THETA(3)))
IF (GENOTYPE1) THEN GENE = G1
IF (GENOTYPE2) THEN GENE = G2
IF (GENOTYPEUNK) THEN GENE = G1*FRA+G2*(1-FRA)

You can then compare FRA to the expected genotypic distribution in the 
population.

Thanks,

Bill

-Original Message-
From: owner-nmus...@globomaxnm.com [mailto:owner-nmus...@globomaxnm.com] On 
Behalf Of Leonid Gibiansky
Sent: Wednesday, November 19, 2014 10:11 AM
To: Jeroen Elassaiss-Schaap; 이소정
Cc: nmusers@globomaxnm.com
Subject: Re: [NMusers] Genotype data missing in some individuals

I would do mixture model only if there is a very large -several folds- 
difference in PK parameters for two genotypes. If the difference is comparable 
with the inter-subject variability within the genotype, I would introduce 
category "missing" to remove the effect of those subjects on covariate effect 
estimate. So if the genotype is binary (YES/NO), you introduce the new third 
level "missing", work with it as with the 3-level categorical covariate, and 
report the difference between NO and YES as the genotype effect on PK. As a 
check for consistency, you may want to check whether the estimate of the PK 
parameter for "missing" level is somewhere between the estimates for "NO" and 
"YES" levels, closer to the value for the level with higher prevalence in your 
dataset.
Regards,
Leonid

--
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web:www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel:(301) 767 5566



On 11/19/2014 6:16 AM, Jeroen Elassaiss-Schaap wrote:
> Dear SoJeong,
>
> First you might want to answer the question whether that phenotype is 
> indeed important in your dataset. With the initial popPK model you 
> could plot posthoc clearance against bodyweight and/or inspect the 
> posthocs of clearance for evidence of multiple peaks in your 
> distribution. You also may see the impact of phenotype in stratified 
> concentration versus time plots. Depending on the dataset, with its 
> sampling scheme, number of subjects (perhaps a low number) and 
> distribution across age, it could be masked.
>
> If the impact is clear however, it might be benificial to try to 
> include the subjects wih missing genotype. With a clear effect, you 
> might be able to develop a mixture model. The mixture  approach would 
> describe the different populations in your dataset corresponding to 
> the different phenotypes. The genotype would than inform the mixture 
> as a covariate - the missing information would fall back to the pure 
> mixture approach. As a warning, this approach is quite difficult. I 
> would advise you to read up on the nonmem guides ($MIX) on this and 
> look in the literature for examples - the Karlsson group has published 
> about it, most recently this one (it contains code):
> http://link.springer.com/article/10.1208/s12248-009-9093-4. A search 
> in the literature gives you additional background such as 
> http://www.page-meeting.org/pdf_assets/9595-PAGE2007_3.pdf and 
> http://link.springer.com/article/10.1007/s10928-006-9038-9.
>
> If the impact is not clear, a more empirical approach might be called 
> for, in this case a subset analysis, i.e. where you exclude the 
> missing subjects, of the covariate relationship might be all that you 
> could achieve. If there is no impact at all, you do not need the 
> genotype of course.
>
> Hope this helps!
>
> Best regards,
>
> Jeroen
>
> http://pd-value.com
> jer...@pd-value.com
> @PD_value
> +31 6 23118438
> -- More value out of your data!
>
> On Nov 19, 2014, at 7:57 AM, "이소정"  > wrote:
>
> Dear all,
>
> I’ve analyzed a tacrolimus PopPK in pediatric patients.
>
> As you know, CYP3A5 genotype can change the tacrolimus PK
> significantly, 3A5 genotyping was performed in the study,
>
> however, in 20% of the subjects, the genotype data was missed.
>
> Then, how can I reflect the CYP3A5 genotype effect to the tacrolimus
> population model appropriately?
>
> Is there any solution?
>
> Best regards,
>
> SoJeong Yi
>

RE: [NMusers] Genotype data missing in some individuals

[Mats Karlsson]

Hi,

I would use:
IF (GENOTYPE.EQ.1)  GENE = THETA(1)
IF (GENOTYPE.EQ.2)  GENE = THETA(2) 
IF (GENOTYPE.EQ.-99.AND.MIXNUM.EQ.1)   GENE = THETA(1) 
IF (GENOTYPE.EQ.-99.AND.MIXNUM.EQ.2)   GENE = THETA(2)
$MIX
P(1)=THETA(3)
P(2)=1-P(1)
..

To handle THETA(3) there are different options
If I believe that missingness is completely at random  (MCAR):
THETA(3) can be fixed to the frequency of GENOTYPE=1 in the population you are 
studying if it is known what this frequency is.
If it is not known, I would fix it to the fraction of GENOTYPE=1 in your 
sample. If I was really ambitious, I would take into account that your sample 
is small and therefore it may not perfectly reflect the proportion in your 
population. If so you could use the prior functionality.
If you believe missingness is missing at random (MAR) another approach could be 
implemented. [MAR in this case could be that there are more missing of one 
ethnic group than another, but you know the ethnicity of everyone.] You would 
only modify one line in the code above:
$MIX
P(1)=THETA(3)
IF(ETHNICITY.EQ.2) P(1)= THETA(4)

If you believe that missingness could be not at random (MNAR), for example that 
genotyping failed more often for sujects with true GENO=1, then use the top 
code but estimate THETA(3) would be the appropriate thing to do. 

There are other options too. Two recent articles on this are provided below 
with comparison between methods. Also it describes a multiple imputation 
routine that we recently implemented in PsN.

Comparison of methods for handling missing covariate data.
Johansson ÅM, Karlsson MO.
AAPS J. 2013 Oct;15(4):1232-41. doi: 10.1208/s12248-013-9526-y.

Multiple imputation of missing covariates in NONMEM and evaluation of the 
method's sensitivity to η-shrinkage.
Johansson ÅM, Karlsson MO.
AAPS J. 2013 Oct;15(4):1035-42. doi: 10.1208/s12248-013-9508-0.

Best regards,
Mats



Mats Karlsson, PhD
Professor of Pharmacometrics
 
Dept of Pharmaceutical Biosciences
Faculty of Pharmacy
Uppsala University
Box 591
75124 Uppsala
 
Phone: +46 18 4714105
Fax + 46 18 4714003
www.farmbio.uu.se/research/researchgroups/pharmacometrics/

-Original Message-
From: owner-nmus...@globomaxnm.com [mailto:owner-nmus...@globomaxnm.com] On 
Behalf Of Denney, William S.
Sent: Wednesday, November 19, 2014 6:24 PM
To: Leonid Gibiansky; Jeroen Elassaiss-Schaap; 이소정
Cc: nmusers@globomaxnm.com
Subject: RE: [NMusers] Genotype data missing in some individuals

Hi SoJeong,

I agree with Leonid here on the value of the mixture model.  With potentially 
subtle changes, mixture models can be very difficult.  One way that I've had 
luck previously with a similar approach is to make "unknown genotype" a 
separate category and then to fit a parameter that is fraction "yes" (similar 
to a mixture model, but not specifying a genotype for a subject).  Something 
like:

G1 = THETA(1)
G2 = THETA(2)
FRA = 1/(1+EXP(-THETA(3)))
IF (GENOTYPE1) THEN GENE = G1
IF (GENOTYPE2) THEN GENE = G2
IF (GENOTYPEUNK) THEN GENE = G1*FRA+G2*(1-FRA)

You can then compare FRA to the expected genotypic distribution in the 
population.

Thanks,

Bill

-Original Message-
From: owner-nmus...@globomaxnm.com [mailto:owner-nmus...@globomaxnm.com] On 
Behalf Of Leonid Gibiansky
Sent: Wednesday, November 19, 2014 10:11 AM
To: Jeroen Elassaiss-Schaap; 이소정
Cc: nmusers@globomaxnm.com
Subject: Re: [NMusers] Genotype data missing in some individuals

I would do mixture model only if there is a very large -several folds- 
difference in PK parameters for two genotypes. If the difference is comparable 
with the inter-subject variability within the genotype, I would introduce 
category "missing" to remove the effect of those subjects on covariate effect 
estimate. So if the genotype is binary (YES/NO), you introduce the new third 
level "missing", work with it as with the 3-level categorical covariate, and 
report the difference between NO and YES as the genotype effect on PK. As a 
check for consistency, you may want to check whether the estimate of the PK 
parameter for "missing" level is somewhere between the estimates for "NO" and 
"YES" levels, closer to the value for the level with higher prevalence in your 
dataset.
Regards,
Leonid

--
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web:www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel:(301) 767 5566



On 11/19/2014 6:16 AM, Jeroen Elassaiss-Schaap wrote:
> Dear SoJeong,
>
> First you might want to answer the question whether that phenotype is 
> indeed important in your dataset. With the initial popPK model you 
> could plot posthoc clearance against bodyweight and/or inspect the 
> posthocs of clearance for evidence of multiple peaks in your 
> distribution. You also may see the impact of phenotype in stratified 
> concentration versus time plots. Depending on the dataset, with its 
> sampling scheme, number of subjects (perhaps a low number) and 
> distribution across age, it 

RE: [NMusers] Genotype data missing in some individuals

[Sebastian Frechen]

Hey Bill & SoJeong,

in your suggestion (Bill), you would estimate a fixed effect for the unknown 
genotype weighting between G1 and G2. 
But let's assume that the status "genotype unknown" is completely at random, so 
some are of type 1 and some are of type 2.

Wouldn't it be possible to implement then something like this (not tested, just 
an idea!!):

---
G1 = THETA(1)
G2 = THETA(2)

FRA = PHI(ETA(1))  ; with PHI() as the cdf of the standard normal distribution

IF (GENOTYPE1) THEN GENE = G1
IF (GENOTYPE2) THEN GENE = G2
IF (GENOTYPEUNK) THEN GENE = G1*FRA+G2*(1-FRA)
...
...
$OMEGA 1 FIX
---

So, with fixing the variance of ETA(1) to 1, NONMEM can obtain standard 
normally distributed random effects in the conditional estimation step. 
Evaluating these etas with the cdf of the standard normal distribution (PHI), 
we obtain uniformly distributed effects (between 0 and 1). Hence, for each 
subject with GENOTYPEUNK, we obtain a probability (weight) for G1 and G2, 
respectively, based on the current parameter estimates of the model.

This might be a quasi mixture model approach.


Von: owner-nmus...@globomaxnm.com [owner-nmus...@globomaxnm.com]" im 
Auftrag von "Denney, William S. [william.s.den...@pfizer.com]
Gesendet: Mittwoch, 19. November 2014 18:23
An: Leonid Gibiansky; Jeroen Elassaiss-Schaap; 이소정
Cc: nmusers@globomaxnm.com
Betreff: RE: [NMusers] Genotype data missing in some individuals

Hi SoJeong,

I agree with Leonid here on the value of the mixture model.  With potentially 
subtle changes, mixture models can be very difficult.  One way that I've had 
luck previously with a similar approach is to make "unknown genotype" a 
separate category and then to fit a parameter that is fraction "yes" (similar 
to a mixture model, but not specifying a genotype for a subject).  Something 
like:

G1 = THETA(1)
G2 = THETA(2)
FRA = 1/(1+EXP(-THETA(3)))
IF (GENOTYPE1) THEN GENE = G1
IF (GENOTYPE2) THEN GENE = G2
IF (GENOTYPEUNK) THEN GENE = G1*FRA+G2*(1-FRA)

You can then compare FRA to the expected genotypic distribution in the 
population.

Thanks,

Bill

-Original Message-
From: owner-nmus...@globomaxnm.com [mailto:owner-nmus...@globomaxnm.com] On 
Behalf Of Leonid Gibiansky
Sent: Wednesday, November 19, 2014 10:11 AM
To: Jeroen Elassaiss-Schaap; 이소정
Cc: nmusers@globomaxnm.com
Subject: Re: [NMusers] Genotype data missing in some individuals

I would do mixture model only if there is a very large -several folds- 
difference in PK parameters for two genotypes. If the difference is comparable 
with the inter-subject variability within the genotype, I would introduce 
category "missing" to remove the effect of those subjects on covariate effect 
estimate. So if the genotype is binary (YES/NO), you introduce the new third 
level "missing", work with it as with the 3-level categorical covariate, and 
report the difference between NO and YES as the genotype effect on PK. As a 
check for consistency, you may want to check whether the estimate of the PK 
parameter for "missing" level is somewhere between the estimates for "NO" and 
"YES" levels, closer to the value for the level with higher prevalence in your 
dataset.
Regards,
Leonid

--
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web:www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel:(301) 767 5566



On 11/19/2014 6:16 AM, Jeroen Elassaiss-Schaap wrote:
> Dear SoJeong,
>
> First you might want to answer the question whether that phenotype is
> indeed important in your dataset. With the initial popPK model you
> could plot posthoc clearance against bodyweight and/or inspect the
> posthocs of clearance for evidence of multiple peaks in your
> distribution. You also may see the impact of phenotype in stratified
> concentration versus time plots. Depending on the dataset, with its
> sampling scheme, number of subjects (perhaps a low number) and
> distribution across age, it could be masked.
>
> If the impact is clear however, it might be benificial to try to
> include the subjects wih missing genotype. With a clear effect, you
> might be able to develop a mixture model. The mixture  approach would
> describe the different populations in your dataset corresponding to
> the different phenotypes. The genotype would than inform the mixture
> as a covariate - the missing information would fall back to the pure
> mixture approach. As a warning, this approach is quite difficult. I
> would advise you to read up on the nonmem guides ($MIX) on this and
> look in the literature for examples - the Karlsson group has published
> about it, most recently this one (it contains code):
> http://link.springer.com/article/10.1208/s12248-009-9093-4. A search
> in the literature gives you additional background such as
> http://www.page-meeting.org/pdf_assets/9595-PAGE2007_3.pdf and
> http://link.springer.com/article/10.1007/s10928-006-9038-9.
>
> 

[NMusers] A question about handling large-scale data

[Liudongyang_hotmail]

Hello All Nonmem Users,

 

  I am modeling intra-gastric H+ concentrations as PD biomarker, which
varies from 10e-7 to 10e-1. I log-transformed original data and used "Y =
LOG(IPRED)+EPS(1)" as log-linear error model firstly. The profile could be
simulated well, but when I fitted data, error messages as "rounding error ."
or "numerical difficulty ." showed up. Fitting was terminated generally.
Will anybody share their experiences or tips on this kind of data?

  MANY THANKS IN ADVANCE!

 

Cheers,

Dongyang Liu, Clinical Pharmacologist

Phase I Unit, Clinical Pharmacology Research Center,

Peking Union Medical College Hospital, Beijing, China

M.P.: +86-18610966092

O.P.: +86-10-69158356