[NMusers] qPharmetra is hiring!

[Klaas Prins]

qPharmetra is hiring!

Are you curious, motivated to share your knowledge with colleagues and clients, 
do you care about consistency and reproducibility in work product, and do you 
thrive when working in a collaborative environment? How about an exciting 
opportunity to join our group of enthusiastic pharmacometricians across the US 
and EU? At qPharmetra we are looking to expand our team with appropriate 
candidates! We are looking for skilled pharmacometricians as well as for a Data 
Programmer & PK Analyst for NCA. We prefer candidates with relevant educational 
background (Pharmaceutical Sciences, Medicine, Bioengineering, Statistics, 
etc.), and 5+ years' work experience, and having the skills to use our key 
tools (NONMEM, R, LaTeX, etc.). You might join our US offices in Boston 
Massachusetts or Research Triangle Park, North Carolina, or our European 
offices in Stockholm, Sweden and Nijmegen, the Netherlands. Alternatively, you 
might work from your home. Please get in touch and share with us your 
enthusiasm and motivation to join our ranks! Adding a resume will be highly 
appreciated.

PS: No recruiters please.
PS2: this is a resend of a previous email that had a problem in the signature 
and thus may have not have reached you (blocked) or been sent to you with an 
error message. Apologies for this.

____
Klaas Prins, PhD
Vice President EU

+31 618 802 292 (M)
+31 248 457 290 (O)
klaas.pr...@qpharmetra.com<mailto:klaas.pr...@qpharmetra.com> | 
www.qpharmetra.com<http://www.qpharmetra.com/>

[NMusers] qPharmetra is hiring!

[Klaas Prins]

qPharmetra is hiring!

Are you curious, motivated to share your knowledge with colleagues and clients, 
do you care about consistency and reproducibility in work product, and do you 
thrive when working in a collaborative environment? How about an exciting 
opportunity to join our group of enthusiastic pharmacometricians across the US 
and EU? At qPharmetra we are looking to expand our team with appropriate 
candidates! We are looking for skilled pharmacometricians as well as for a Data 
Programmer & PK Analyst for NCA. We prefer candidates with relevant educational 
background (Pharmaceutical Sciences, Medicine, Bioengineering, Statistics, 
etc.), and 5+ years' work experience, and having the skills to use our key 
tools (NONMEM, R, LaTeX, etc.). You might join our US offices in Boston 
Massachusetts or Research Triangle Park, North Carolina, or our European 
offices in Stockholm, Sweden and Nijmegen, the Netherlands. Alternatively, you 
might work from your home. Please get in touch and share with us your 
enthusiasm and motivation to join our ranks! Adding a resume will be highly 
appreciated.

PS: No recruiters please.

____
Klaas Prins, PhD
Vice President EU
qPharmetra, LLC

+31 618 802 292 (M)
+31 248 457 290 (O)
klaas.pr...@qpharmetra.com<mailto:kevin.dyks...@qpharmetra.com> | 
www.qpharmetra.com<http://www.qpharmetra.com/>

RE: [NMusers] BAYES Estimation: OBJECTIVE FUNCTION IS INFINITE

[Klaas Prins]

Hi Xinting,

I frequently use METH=BAYES and have had the same observation at various 
occasions. First let me say that BAYES a wonderful estimation engine that has 
helped me to model PK, PKPD and odd-type data in cases where FOCE failed, even 
where SAEM and IMP(MAP) failed (recently shown at PAGE). In all situations of 
infinite OFV I managed to get it to run by making sure the search would not 
lead to extremely large or extremely small ( positive ) numbers. For example if 
you are looking for a power estimate and you have some idea about it going to 
be in the order of 1, then ensure that your safety valves in the code prevent 
it from going close to zero or larger than e.g. 3 (arbitrary number). While the 
estimate at extreme values may not be problematic per se, one could think that 
in combination with the other parameters (I believe BAYES considers all 
parameters ‘completely free to choose’ within the initial estimate bounds 
during estimation) completely weird predictions could lead to the OFV being 
blown up. If I read about your problem being high-dimensional, there may be 
many relationships in your model at large to make the OFV go poof. One argument 
against directing the parameters in a certain direction is that it may 
introduce bias, so it comes at a certain cost. Furthermore I can recommend to 
look into the bayes.ext file you’re writing to disk and you may see at what 
parameter combinations OFV goes bazurk; that may inform you further on extreme 
values.

Apparently BAYES is more flexible in the search which provides combinations of 
parameters that will send the OFV into the wild. Perhaps Bob Bauer can add some 
insights here – I know too little about what happens under the hood.

By the way, why are you considering BAYES when you get SAEM and IMP to run just 
fine? Why not use SAEM results and run IMP with EONLY=1 afterwards?

Hope this helps,
Klaas Prins


Klaas Prins, PhD
Vice President EU
[QPharmetra_noTag_FA little.png]
+31 618 802 292 (M)
+31 248 457 290 (O)
klaas.pr...@qpharmetra.com<mailto:kevin.dyks...@qpharmetra.com>

This e-mail communication is confidential and is intended only for the 
individual(s) or entity named above and others who have been specifically 
authorized to receive it. If you are not the intended recipient, please do not 
read, copy, use or disclose the contents of this communication to others. 
Please notify the sender that you have received this e-mail in error by 
replying to the e-mail or by telephoning +31 618 802 292. Please then delete 
the e-mail and any copies of it. Thank you.



From: owner-nmus...@globomaxnm.com [mailto:owner-nmus...@globomaxnm.com] On 
Behalf Of Xinting Wang
Sent: Friday, June 12, 2015 4:06 AM
To: nmusers@globomaxnm.com
Subject: [NMusers] BAYES Estimation: OBJECTIVE FUNCTION IS INFINITE

Dear All,
I am using a cascade of the EM estimation methods implemented in Nonmem 7.2 to 
solve a high-dimension PK/PD problem. The estimation chain is coded as below:

$EST MET=SAEM INTER NBURN=2000 NITER=1000 PRINT=5 NOABORT NOPRIOR=0 ISAMPLE=10 
SIGL=6 CTYPE=3 IACCEPT=0.4 CONSTRAIN=1 SEED=145612 FILE=Saem.ext

$EST MET=IMP INTER NITER=2000 ISAMPLE=1000 PRINT=5 SEED=581987 SIGL=6, CTYPE=3 
MAPITER=0 FILE=Imp.ext

$EST MET=BAYES INTER NBURN=3000 NITER=25000 PRINT=10 NOABORT NOPRIOR=0 FNLETA=0 
ISAMPLE_M1=2 ISAMPLE_M2=2 ISAMPLE_M3=2 IACCEPT=0.4 SEED=231457 FILE=Bayes.ext

​The first two estimation methods run successfuly with a convergence, but when 
it comes to BAYES, an error message always occurs:

​ OBJECTIVE FUNCTION IS INFINITE. PROBLEM ENDED
1THERE ARE ERROR MESSAGES IN FILE PRDERR

​I put safety brakes in the code to prevent ​0**power or square root of 
negative values, but the problem still occured. Setting lower and upper bound 
of the initial estimation, but it did not help either.

I am wondering if anyone here encountered this issue previously? It seems quite 
strange to me that SAEM and IMP could succeed, but BAYES fails. Could you 
please let me know of the possible reasons behind this? Thanks very much.
Best Regards

--
Xinting

RE: [NMusers] variation of time values in output data

[Klaas Prins]

If you subtract (DATE-1) * 24 from TIME in the output file it will match TIME 
in input file.
NONMEM reads DATE as a DATE data item, that’s why. I suspect you won’t have 
this if you choose another column name for DATE (e.g. DAY). Note that this may 
also have an effect on the estimation process and thus your parameter estimates 
may change as well (for the better).

Best regards,
Klaas


From: owner-nmus...@globomaxnm.com [mailto:owner-nmus...@globomaxnm.com] On 
Behalf Of Diane-Charlotte Imbs
Sent: Tuesday, April 21, 2015 11:47 AM
To: Leonid Gibiansky
Cc: nmusers@globomaxnm.com
Subject: Re: [NMusers] variation of time values in output data

Thank you, Leonid, for your answer but you can see (below) it’s not a rounding 
issue ; there are udge variations of time ?!!!

INPUT FILE :

ID   DATE TIME  DV  AMT  RATE 
MDV  EVID
3 1 00   2000-2 
 1 1
3 1 2.6 1060   .  . 
  0 0
3 1 170.117   331 .  .  
 0 0
3 2 170.117   . 2000   -2   
   1 1
3 2 172.117   1220   .  .   
0 0
3 2 326.333   220 . .   
0 0
3 3 326.55 .   2000-2   
  1 1
3 3 328.617   964 . .   
0 0
3 3 504.833   1260   . .
   0 0

OUTPUT FILE :

ID   DATE TIME  IPREDCL   D1
  V1  IWRES
3 1 0   0 1.7711   
113.17   35.2170
3 1 2.61223  1.7711   
113.17   35.217 -0.13328
3 1 170.12   567.29   1.7711   113.17   
35.217 -0.41653
3 2 194.12   169.68   1.7711   112.49   
35.217 -1
3 2 196.12   1114  1.7711   112.49  
 35.217  0.09514
3 2 350.33   1110  1.7711   112.49  
 35.217 -0.80181
3 3 374.55   328.36   1.7711   140.63   
35.217 -1
3 3 376.62   1089.8   1.7711   140.63   
35.217 -0.11542
3 3 552.83   1208.1   1.7711   140.63   
35.217 0.04297



Loïc FIEVET
Interne IPR en pharmacocinétique

Institut Universitaire du Cancer Toulouse - Oncopole
1 avenue Irène Joliot-Curie
31059 TOULOUSE Cedex 9









2015-04-20 19:21 GMT+02:00 Leonid Gibiansky 
mailto:lgibian...@quantpharm.com>>:
it could be rounding issue when the precision of the TIME in the data file 
exceeds the precision of the TIME in the output of the table. You could 
increase precision of the output using format statement (see manual). Other 
than that, NONMEM should not change TIME variable.

Leonid

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



On 4/20/2015 12:12 PM, Diane-Charlotte Imbs wrote:
Dear nonmem users,

We used the option “-2 RATE” of NONMEM (actual rate of infusion is
unknown): within the control stream:

D1=THETA(1)

Moreover, since several infusions are given to the same patients, we
have added an inter-occasion variability on D1 (see code below)*

We obtained nice fit BUT in the OUTPUT DATA: TIME values corresponding
to blood samples (and DV) are not identical to the actual TIME values of
the INPUT data !?

It seems that NONMEM program adjusts rate values, their interoccasion
variability, and sampling times to have a good fit

*How to fix TIME value corresponding to blood samples ? Indeed, we do
not know the actual infusion rate but we do know the sampling times.*

**

Can anyone tell me if I chose the right model or help me resolving this?

Thanks in advance,

*$PK

OCC1=0

OCC2=0

IF(DATE.EQ.1) OCC1=1

IF(DATE.EQ.2) OCC2=1

D1=THETA(1)*EXP(ETA(3)+ETA(4)*OCC1+ETA(5)*OCC2)

$THETA (0,168,);D1

$OMEGA .1 ; iiv d1

$OMEGA BLOCK(1) .01 ; iov ETA4 OCC1

$OMEGA BLOCK(1) SAME ; iov ETA5 OCC2


*Loïc FIEVET*

*/Interne IPR en pharmacocinétique (1^er semestre)/*

*Institut Universitaire du Cancer Toulouse - Oncopole*

1 avenue Irène Joliot-Curie

31059 TOULOUSE Cedex 9

Re: [NMusers] Question about interoccation variability

[Klaas Prins]

I think I would try IOV on F1 first before putting it on KA like Jean suggested 
but above all I think there is an essential element missing in your code. IOV 
is variability between occasions on top of inter individual variability.

So do something like: 
F1=1
F1=THETA(1)*EXP(ETA(6)+BOVKA)
$OMEGA 0.25 ; IIV F1

Furthermore, there may be other elements contributing to the inability to 
predict Cmax well, such as more complex absorption features. I think we lack 
info to comment on that.

HTH, Klaas 


On 10 jul. 2012, at 23:56, "Lavigne, Jean"  wrote:

> KA=THETA(1)*EXP(BOVKA)