Hi
I'm an undergraduate student, now using NONMEM analyse PK data(adv2tr2 model).
But it keep showing ELAPSED TIME MAY NOT BE NEGATIVE.
THE DV DATA ITEM IS POSITIVE, BUT THE MDV DATA ITEM IS 1.
I've already added the MDV item, but still no change.
Is it about data formulation problems or there is other issues?
Thank you
Hello Tsai
The data file of NONMEM has specific structural requirements to work, you should include a dosing record (MDV=1) and an observation record (MDV=0). If you have a DV as you stated your MDV should be 0 not 1.
Also check on your times format, a common error resulting in negative times is inconsistent formatting of the times between different records.
Hope this helps
Kindest regards
Dr. Mariam Abdel Jalil
The message is that elapsed TIME (!) may not be negative, so why look into DV and MDV but not TIME?
TIME is indeed expected to not be negative, see this discussion:
http://www.cognigencorp.com/nonmem/current/2013-December/4891.html
A common way of tracking down the problem is by successively removing or commenting out records (rows) until the error disappears. That way you know where the trouble originates.
It would be reproducible if you posted the data set in question and even the control stream if you can.
Dear Dr Krause and Jalil
Thanks for your teaching!
It's a little bit hard for me to check the control stream and these bugs.
The record format looks the same.
I've all used the hh;mm format to edit the data, but there are 3 missing times.
Is that a problem for running this model?
I've put partial data and model on.
I'll be appreciate for you checking them if there's free time.
Thank you
Dear Tsai
NONMEM identifies the entered data from your $input in the control stream. Your fourth data item in the control stream is amount. So NONMEM is expecting to find numbers not times in your fourth column of the data.
Other noted issues:
Try to find what the missing times are, do not leave blanks in your dataset
Your date format is confusing: You can use calendar dates or integer numbers reflecting study days (but they should be ordered chronologically).
Finally please refere to
NONMEM Users Guide Part V - Introductory Guide - Chapter 6
as it might be useful to understand the use of MDV and EVID and general formatting of the datafile and control file.
Hope this helps
Kindest regards
Dr. Mariam Abdel Jalil
This next problem looks obvious:
what you specify as headers in the control stream,
INPUT C ID TIMEO AMT DV EVID MDV TIME ;RATE etc.
does not match the headers in the data set,
C ID DATE TIMEO AMT DV EVID MDV SEX PCA TBW CHF
To preempt the next question, nonmem does not appreciate missing data (empty fields), as pointed out by Mariam.
- Badges
- Science topic
- Similar topics
- Chemistry
- Pharmacology
- Pharmaceuticals
- Drugs