[NMusers] NONMEM error
Colleagues Yesterday, I asked the group for suggestions to deal with a NONMEM error that appeared to result from a lengthy interval between doses causing NONMEM to abort. I received many suggestions. The one that I implemented successfully was to add records with EVID=3 2.0 days after the record at the beginning of the interval (suggested by Bill Denney and Jeroen Elassaiss-Schapp via email and Steve Shafer via phone). For example, if there were dose records at 0, 30, and 60-90 days, I added records at 2.0 and 32.0 days. Problem solved. Thanks for the prompt response from the community. Dennis Dennis Fisher MD P < (The "P Less Than" Company) Phone / Fax: 1-866-PLessThan (1-866-753-7784) www.PLessThan.com
RE: [NMusers] NONMEM error
Hi Dennis, One other idea that can help if all measures are BQL is to use a reset record or reset and dose (EVID 3 or 4) on the subsequent doses. That way, the integrator will be restarted for those participants and may help. Thanks, Bill -Original Message- From: owner-nmus...@globomaxnm.com On Behalf Of Leonid Gibiansky Sent: Thursday, December 5, 2024 3:14 PM To: Dennis Fisher ; nmusers@globomaxnm.com Subject: Re: [NMusers] NONMEM error Are there any measurements that are not BQLs and that are more than 24 hrs from the last dose? What is the residual error ? Combined error (rather than proportional or exponential) may help. Random effect on residual error is very helpful to identify aberrant subjects/observations. $ABBREV PROTECT helps sometimes. NOHABORT (with H) can help but it is dangerous, so it can be used for debugging, but not for the final models. I would look on the data with large WRES (after the FO run) to see whether any of the observations are so far off that Nonmem cannot fit them and crashes. Thank you Leonid On 12/5/2024 2:33 PM, Dennis Fisher wrote: > Colleagues > > I have encountered a problem during NONMEM execution. > > Most subjects are dosed daily but a small number of subjects have > large intervals (> 30 days) between successive doses. > For subjects dosed daily, concentrations 24 hours after a dose are BQL > and there is no evidence of accumulation. > > For the problem subjects, NONMEM either: > 1. aborts (despite a NOABORT option) with an error message at the > first dose after the lengthy gap 2. or sends the following error > message: > > 0PRED EXIT CODE = 1 > 0INDIVIDUAL NO. 103 ID= 3.110400E+04 > (WITHIN-INDIVIDUAL) DATA REC NO. 56 > THETA= >1.66E+02 1.15E-01 2.95E+01 1.81E+01 4.33E+01 3.01E+02 > 8.26E+00 0.00E+00 1.00E+00 0.00E+00 >0.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 > 1.00E+00 > OCCURS DURING SEARCH FOR ETA AT INITIAL VALUE, ETA=0 > NUMERICAL DIFFICULTIES WITH INTEGRATION ROUTINE. > MAXIMUM NO. OF EVALUATIONS OF DIFFERENTIAL EQUATIONS, 100, > EXCEEDED. > > The error occurs with METHOD=1 but NOT with METHOD=0. > > For one subject in whom doses were administered on Days 1, 30, and 60 > (followed by daily dosing), I deleted the first two dosing records and > the error no longer occurs with METHOD=1. I speculate that the error > might relate to the predicted concentration becoming Infinitesimally > small after a lengthy interval between doses, However, I have no idea > if this is the explanation. And I am reluctant to change the dataset > so I seek other solutions. > > Any ideas as to an explanation or a work-around? > > Dennis > > > Dennis Fisher MD > P < (The "P Less Than" Company) > Phone / Fax: 1-866-PLessThan (1-866-753-7784) www.PLessThan.com >
Re: [NMusers] NONMEM error
Dear Dennis, " NUMERICAL DIFFICULTIES WITH INTEGRATION ROUTINE. MAXIMUM NO. OF EVALUATIONS OF DIFFERENTIAL EQUATIONS, 100, EXCEEDED. " This seem to indicate that is due to large gap from the first dose time to the next, causing the difficulty in numerical integration. For these subjects, can you try to add few dummy records with EVID=2 to resolve this problem. Best regards, Sam On 12/5/2024 11:21 AM, Dennis Fisher wrote: Colleagues I have encountered a problem during NONMEM execution. Most subjects are dosed daily but a small number of subjects have large intervals (> 30 days) between successive doses. For subjects dosed daily, concentrations 24 hours after a dose are BQL and there is no evidence of accumulation. For the problem subjects, NONMEM either: 1. aborts (despite a NOABORT option) with an error message at the first dose after the lengthy gap 2. or sends the following error message: 0PRED EXIT CODE = 1 0INDIVIDUAL NO. 103 ID= 3.110400E+04 (WITHIN-INDIVIDUAL) DATA REC NO. 56 THETA= 1.66E+02 1.15E-01 2.95E+01 1.81E+01 4.33E+01 3.01E+02 8.26E+00 0.00E+00 1.00E+00 0.00E+00 0.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 OCCURS DURING SEARCH FOR ETA AT INITIAL VALUE, ETA=0 NUMERICAL DIFFICULTIES WITH INTEGRATION ROUTINE. MAXIMUM NO. OF EVALUATIONS OF DIFFERENTIAL EQUATIONS, 100, EXCEEDED. The error occurs with METHOD=1 but NOT with METHOD=0. For one subject in whom doses were administered on Days 1, 30, and 60 (followed by daily dosing), I deleted the first two dosing records and the error no longer occurs with METHOD=1. I speculate that the error might relate to the predicted concentration becoming Infinitesimally small after a lengthy interval between doses, However, I have no idea if this is the explanation. And I am reluctant to change the dataset so I seek other solutions. Any ideas as to an explanation or a work-around? Dennis Dennis Fisher MD P < (The "P Less Than" Company) Phone / Fax: 1-866-PLessThan (1-866-753-7784) www.PLessThan.com
Re: [NMusers] NONMEM error
Are there any measurements that are not BQLs and that are more than 24 hrs from the last dose? What is the residual error ? Combined error (rather than proportional or exponential) may help. Random effect on residual error is very helpful to identify aberrant subjects/observations. $ABBREV PROTECT helps sometimes. NOHABORT (with H) can help but it is dangerous, so it can be used for debugging, but not for the final models. I would look on the data with large WRES (after the FO run) to see whether any of the observations are so far off that Nonmem cannot fit them and crashes. Thank you Leonid On 12/5/2024 2:33 PM, Dennis Fisher wrote: Colleagues I have encountered a problem during NONMEM execution. Most subjects are dosed daily but a small number of subjects have large intervals (> 30 days) between successive doses. For subjects dosed daily, concentrations 24 hours after a dose are BQL and there is no evidence of accumulation. For the problem subjects, NONMEM either: 1. aborts (despite a NOABORT option) with an error message at the first dose after the lengthy gap 2. or sends the following error message: 0PRED EXIT CODE = 1 0INDIVIDUAL NO. 103 ID= 3.110400E+04 (WITHIN-INDIVIDUAL) DATA REC NO. 56 THETA= 1.66E+02 1.15E-01 2.95E+01 1.81E+01 4.33E+01 3.01E+02 8.26E+00 0.00E+00 1.00E+00 0.00E+00 0.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 OCCURS DURING SEARCH FOR ETA AT INITIAL VALUE, ETA=0 NUMERICAL DIFFICULTIES WITH INTEGRATION ROUTINE. MAXIMUM NO. OF EVALUATIONS OF DIFFERENTIAL EQUATIONS, 100, EXCEEDED. The error occurs with METHOD=1 but NOT with METHOD=0. For one subject in whom doses were administered on Days 1, 30, and 60 (followed by daily dosing), I deleted the first two dosing records and the error no longer occurs with METHOD=1. I speculate that the error might relate to the predicted concentration becoming Infinitesimally small after a lengthy interval between doses, However, I have no idea if this is the explanation. And I am reluctant to change the dataset so I seek other solutions. Any ideas as to an explanation or a work-around? Dennis Dennis Fisher MD P < (The "P Less Than" Company) Phone / Fax: 1-866-PLessThan (1-866-753-7784) www.PLessThan.com
Re: [NMusers] NONMEM error
Hi Dennis,You could try any of the following to prevent the suspected numerical issue:- insert $ABBR PROTECT- set ONLY OBSERVATIONS in the error block - insert a reset event (EVID=4) after the 30-day hiatusHope this helps,Jeroenhttp://pd-value.comjer...@pd-value.com@PD_value+31 6 23118438-- More value out of your data!Op 5 dec 2024 om 20:45 heeft Dennis Fisher het volgende geschreven:ColleaguesI have encountered a problem during NONMEM execution. Most subjects are dosed daily but a small number of subjects have large intervals (> 30 days) between successive doses. For subjects dosed daily, concentrations 24 hours after a dose are BQL and there is no evidence of accumulation. For the problem subjects, NONMEM either:1. aborts (despite a NOABORT option) with an error message at the first dose after the lengthy gap2. or sends the following error message:0PRED EXIT CODE = 10INDIVIDUAL NO. 103 ID= 3.110400E+04 (WITHIN-INDIVIDUAL) DATA REC NO. 56 THETA= 1.66E+02 1.15E-01 2.95E+01 1.81E+01 4.33E+01 3.01E+02 8.26E+00 0.00E+00 1.00E+00 0.00E+00 0.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 OCCURS DURING SEARCH FOR ETA AT INITIAL VALUE, ETA=0 NUMERICAL DIFFICULTIES WITH INTEGRATION ROUTINE. MAXIMUM NO. OF EVALUATIONS OF DIFFERENTIAL EQUATIONS, 100, EXCEEDED. The error occurs with METHOD=1 but NOT with METHOD=0.For one subject in whom doses were administered on Days 1, 30, and 60 (followed by daily dosing), I deleted the first two dosing records and the error no longer occurs with METHOD=1. I speculate that the error might relate to the predicted concentration becoming Infinitesimally small after a lengthy interval between doses, However, I have no idea if this is the explanation. And I am reluctant to change the dataset so I seek other solutions.Any ideas as to an explanation or a work-around?Dennis Dennis Fisher MDP < (The "P Less Than" Company)Phone / Fax: 1-866-PLessThan (1-866-753-7784)www.PLessThan.com
[NMusers] NONMEM error
Colleagues I have encountered a problem during NONMEM execution. Most subjects are dosed daily but a small number of subjects have large intervals (> 30 days) between successive doses. For subjects dosed daily, concentrations 24 hours after a dose are BQL and there is no evidence of accumulation. For the problem subjects, NONMEM either: 1. aborts (despite a NOABORT option) with an error message at the first dose after the lengthy gap 2. or sends the following error message: 0PRED EXIT CODE = 1 0INDIVIDUAL NO. 103 ID= 3.110400E+04 (WITHIN-INDIVIDUAL) DATA REC NO. 56 THETA= 1.66E+02 1.15E-01 2.95E+01 1.81E+01 4.33E+01 3.01E+02 8.26E+00 0.00E+00 1.00E+00 0.00E+00 0.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 OCCURS DURING SEARCH FOR ETA AT INITIAL VALUE, ETA=0 NUMERICAL DIFFICULTIES WITH INTEGRATION ROUTINE. MAXIMUM NO. OF EVALUATIONS OF DIFFERENTIAL EQUATIONS, 100, EXCEEDED. The error occurs with METHOD=1 but NOT with METHOD=0. For one subject in whom doses were administered on Days 1, 30, and 60 (followed by daily dosing), I deleted the first two dosing records and the error no longer occurs with METHOD=1. I speculate that the error might relate to the predicted concentration becoming Infinitesimally small after a lengthy interval between doses, However, I have no idea if this is the explanation. And I am reluctant to change the dataset so I seek other solutions. Any ideas as to an explanation or a work-around? Dennis Dennis Fisher MD P < (The "P Less Than" Company) Phone / Fax: 1-866-PLessThan (1-866-753-7784) www.PLessThan.com
[NMusers] NONMEM error
Colleagues I have encountered a problem during NONMEM execution. Most subjects are dosed daily but a small number of subjects have large intervals (> 30 days) between successive doses. For subjects dosed daily, concentrations 24 hours after a dose are BQL and there is no evidence of accumulation. For the problem subjects, NONMEM either: 1. aborts (despite a NOABORT option) with an error message at the first dose after the lengthy gap 2. or sends the following error message: 0PRED EXIT CODE = 1 0INDIVIDUAL NO. 103 ID= 3.110400E+04 (WITHIN-INDIVIDUAL) DATA REC NO. 56 THETA= 1.66E+02 1.15E-01 2.95E+01 1.81E+01 4.33E+01 3.01E+02 8.26E+00 0.00E+00 1.00E+00 0.00E+00 0.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 1.00E+00 OCCURS DURING SEARCH FOR ETA AT INITIAL VALUE, ETA=0 NUMERICAL DIFFICULTIES WITH INTEGRATION ROUTINE. MAXIMUM NO. OF EVALUATIONS OF DIFFERENTIAL EQUATIONS, 100, EXCEEDED. The error occurs with METHOD=1 but NOT with METHOD=0. For one subject in whom doses were administered on Days 1, 30, and 60 (followed by daily dosing), I deleted the first two dosing records and the error no longer occurs with METHOD=1. I speculate that the error might relate to the predicted concentration becoming Infinitesimally small after a lengthy interval between doses, However, I have no idea if this is the explanation. And I am reluctant to change the dataset so I seek other solutions. Any ideas as to an explanation or a work-around? Dennis Dennis Fisher MD P < (The "P Less Than" Company) Phone / Fax: 1-866-PLessThan (1-866-753-7784) www.PLessThan.com
[NMusers] NONMEM Error on SAEM but not FOCEI
Clearly a bug in the $ABBR PROTECT processor. The NMTRAN interpreter removes the *EXP(0) (because it does not modify the equation), which messes up $ABBR PROTECT process. This occurs only with EXP(0), but not EXP(1) EXP(2), etc. Work-arounds are: KA = TVKA *EXP(-0) or ZERO=0.0 KA = TVKA *EXP(ZERO) Robert J. Bauer, Ph.D. Senior Director Pharmacometrics R&D ICON Early Phase 731 Arbor way, suite 100 Blue Bell, PA 19422 Office: (215) 616-6428 Mobile: (925) 286-0769 robert.ba...@iconplc.com<mailto:robert.ba...@iconplc.com> www.iconplc.com<http://www.iconplc.com/> From: owner-nmus...@globomaxnm.com<mailto:owner-nmus...@globomaxnm.com> mailto:owner-nmus...@globomaxnm.com>> On Behalf Of Mark Sale Sent: Friday, January 26, 2024 10:59 AM To: Leonid Gibiansky mailto:lgibian...@quantpharm.com>>; Jeroen Elassaiss-Schaap (PD-value B.V.) mailto:jer...@pd-value.com>>; Elashkar, Omar I. mailto:omar.elash...@ufl.edu>>; nmusers@globomaxnm.com<mailto:nmusers@globomaxnm.com> Subject: [EXTERNAL] RE: [NMusers] NONMEM Error on SAEM but not FOCEI An interesting undocumented feature of PROTECT. If you have EXP(0), e.g., KA = TVKA *EXP(0) The fortran translation in this case apparently is: KA=TVKA*PEXP(*)B92 Which then throws a syntax error. Doesn't happen without $ABBR PROTECT Not sure if that counts as a bug, but does prohibit using PROTECT in automated methods to run models with/without BSV, or at least makes it a little more difficult. Mark Sale M.D. Vice President Integrated Drug Development mark.s...@certara.com<mailto:mark.s...@certara.com> Remote-Forestville CA Office Hours 9 AM - 5 PM Eastern Time +1 302-516-1684 www.certara.com<http://www.certara.com> ICON plc made the following annotations. -- This e-mail transmission may contain confidential or legally privileged information that is intended only for the individual or entity named in the e-mail address. If you are not the intended recipient, you are hereby notified that any disclosure, copying, distribution, or reliance upon the contents of this e-mail is strictly prohibited. If you have received this e-mail transmission in error, please reply to the sender, so that ICON plc can arrange for proper delivery, and then please delete the message. Thank You, ICON plc South County Business Park Leopardstown Dublin 18 Ireland Registered number: 145835
RE: [NMusers] NONMEM Error on SAEM but not FOCEI
An interesting undocumented feature of PROTECT. If you have EXP(0), e.g., KA= TVKA *EXP(0) The fortran translation in this case apparently is: KA=TVKA*PEXP(*)B92 Which then throws a syntax error. Doesn't happen without $ABBR PROTECT Not sure if that counts as a bug, but does prohibit using PROTECT in automated methods to run models with/without BSV, or at least makes it a little more difficult. Mark Sale M.D. Vice President Integrated Drug Development mark.s...@certara.com Remote-Forestville CA Office Hours 9 AM - 5 PM Eastern Time +1 302-516-1684 www.certara.com -Original Message- From: owner-nmus...@globomaxnm.com On Behalf Of Leonid Gibiansky Sent: Friday, January 26, 2024 7:42 AM To: Jeroen Elassaiss-Schaap (PD-value B.V.) ; Elashkar, Omar I. ; nmusers@globomaxnm.com Subject: Re: [NMusers] NONMEM Error on SAEM but not FOCEI CAUTION: This email originated from outside of Certara. Do not click links or open attachments unless you recognize the sender and know the content is safe. This part is easy to solve: if you do it in log space, Q will always be positive, no PROTECT is needed (and I am not sure it is relevant in the case when Q is going to be negative). On 1/26/2024 10:27 AM, Jeroen Elassaiss-Schaap (PD-value B.V.) wrote: > Dear Omar, > > > What you describe is not that strange. It can happen because SAEM does > random sampling whereas FOCE-I uses a directed search. Therefore > extreme values can occur with SAEM. Please refer to the section I60 of > user guide for v7+ "Stable Routines for Estimation Methods and > Automated Protection Against Floating Point Exceptions" for more > detail, but you could start with including > > $ABBR PROTECT > > which probably is adequate in your case. > > > Hope this helps, > > Jeroen > > > https://nam11.safelinks.protection.outlook.com/?url=http%3A%2F%2Fpd-va > lue.com%2F&data=05%7C02%7CMark.sale%40Certara.com%7Cecb21da5ca6e4a50d8 > b308dc1e865d9a%7C7287abd30220456e98514352bae208c9%7C1%7C0%7C6384188096 > 75762447%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiL > CJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C0%7C%7C%7C&sdata=4uyLVZH1Zx4EWEQws7xn4 > %2FKR%2BeFIro5ul10iI6LMUM4%3D&reserved=0 > jer...@pd-value.com > @PD_value > +31 6 23118438 > -- More value out of your data! > > On 26-01-2024 16:01, Elashkar, Omar I. wrote: >> Hello NMusers, >> >> I have created a model that fit ok with FOCE+I. When I try to fit >> with SAEM, I get the following error: >> >> """ >> ERROR IN TRANS4 ROUTINE: Q IS ZERO >> Elapsed estimation time in seconds: 0.00 >> INDIVIDUAL OBJECTIVE FUNCTION VALUES ARE ALL ZERO. PROBLEM ENDED """ >> >> The data is very small, but the model seems acceptable from bootstrap >> on FOCE+I. Any ideas are appreciated! >> >> Thank you, >> Omar >> >> >> >> *Omar Elashkar, *B.Pharm, MSc >> >> Graduate Assistant >> >> University of Florida | College of Pharmacy >> >> >> > This message (including any attachments) may contain confidential, proprietary, privileged and/or private information. The information is intended to be for the use of the individual or entity designated above. If you are not the intended recipient of this message, please notify the sender immediately, and delete the message and any attachments. Any disclosure, reproduction, distribution or other use of this message or any attachments by an individual or entity other than the intended recipient is prohibited.
Re: [NMusers] NONMEM Error on SAEM but not FOCEI
Depends. I have come across cases with large variances where you get an overflow to zero in SAEM. It is however not very likely with a closed-form PK model. http://pd-value.com jer...@pd-value.com @PD_value +31 6 23118438 -- More value out of your data! On 26-01-2024 16:41, Leonid Gibiansky wrote: This part is easy to solve: if you do it in log space, Q will always be positive, no PROTECT is needed (and I am not sure it is relevant in the case when Q is going to be negative). On 1/26/2024 10:27 AM, Jeroen Elassaiss-Schaap (PD-value B.V.) wrote: Dear Omar, What you describe is not that strange. It can happen because SAEM does random sampling whereas FOCE-I uses a directed search. Therefore extreme values can occur with SAEM. Please refer to the section I60 of user guide for v7+ "Stable Routines for Estimation Methods and Automated Protection Against Floating Point Exceptions" for more detail, but you could start with including $ABBR PROTECT which probably is adequate in your case. Hope this helps, Jeroen http://pd-value.com jer...@pd-value.com @PD_value +31 6 23118438 -- More value out of your data! On 26-01-2024 16:01, Elashkar, Omar I. wrote: Hello NMusers, I have created a model that fit ok with FOCE+I. When I try to fit with SAEM, I get the following error: """ ERROR IN TRANS4 ROUTINE: Q IS ZERO Elapsed estimation time in seconds: 0.00 INDIVIDUAL OBJECTIVE FUNCTION VALUES ARE ALL ZERO. PROBLEM ENDED """ The data is very small, but the model seems acceptable from bootstrap on FOCE+I. Any ideas are appreciated! Thank you, Omar *Omar Elashkar, *B.Pharm, MSc Graduate Assistant University of Florida | College of Pharmacy
Re: [NMusers] NONMEM Error on SAEM but not FOCEI
Dear Omar, What you describe is not that strange. It can happen because SAEM does random sampling whereas FOCE-I uses a directed search. Therefore extreme values can occur with SAEM. Please refer to the section I60 of user guide for v7+ "Stable Routines for Estimation Methods and Automated Protection Against Floating Point Exceptions" for more detail, but you could start with including $ABBR PROTECT which probably is adequate in your case. Hope this helps, Jeroen http://pd-value.com jer...@pd-value.com @PD_value +31 6 23118438 -- More value out of your data! On 26-01-2024 16:01, Elashkar, Omar I. wrote: Hello NMusers, I have created a model that fit ok with FOCE+I. When I try to fit with SAEM, I get the following error: """ ERROR IN TRANS4 ROUTINE: Q IS ZERO Elapsed estimation time in seconds: 0.00 INDIVIDUAL OBJECTIVE FUNCTION VALUES ARE ALL ZERO. PROBLEM ENDED """ The data is very small, but the model seems acceptable from bootstrap on FOCE+I. Any ideas are appreciated! Thank you, Omar *Omar Elashkar, *B.Pharm, MSc Graduate Assistant University of Florida | College of Pharmacy
Re: [NMusers] NONMEM Error on SAEM but not FOCEI
type = "typo", something similar is in the control stream where Q is defined Leonid On 1/26/2024 10:16 AM, Leonid Gibiansky wrote: most likely, some type in the code. Can you provide the control stream? On 1/26/2024 10:01 AM, Elashkar, Omar I. wrote: ERROR IN TRANS4 ROUTINE
Re: [NMusers] NONMEM Error on SAEM but not FOCEI
most likely, some type in the code. Can you provide the control stream? On 1/26/2024 10:01 AM, Elashkar, Omar I. wrote: ERROR IN TRANS4 ROUTINE
[NMusers] NONMEM Error on SAEM but not FOCEI
Hello NMusers, I have created a model that fit ok with FOCE+I. When I try to fit with SAEM, I get the following error: """ ERROR IN TRANS4 ROUTINE: Q IS ZERO Elapsed estimation time in seconds: 0.00 INDIVIDUAL OBJECTIVE FUNCTION VALUES ARE ALL ZERO. PROBLEM ENDED """ The data is very small, but the model seems acceptable from bootstrap on FOCE+I. Any ideas are appreciated! Thank you, Omar Omar Elashkar, B.Pharm, MSc Graduate Assistant University of Florida | College of Pharmacy
