NCA analysis using PKNCA - errors and different results than expected #283
-
|
Rversion = 4.3.2
Note: I am using asnumeric and is.na to make sure my dose, concentration and time are manipulated correctly; I am using read_excel() to read my data and in the beginning I split my data by route creating 2 data frames Thank you so much for your help! |
Beta Was this translation helpful? Give feedback.
Replies: 2 comments 16 replies
-
|
The first issue is the main one. Without fixing it, I'm not sure how you got results of any type. The way that you specified the call to PKNCAconc, you didn't indicate a subject column. Given your description of the study, that is the most likely issue. The way to specify the subject column is to add the subject as the last grouping level similar to the.below call: To get the group mean, you would run For the difference with Phoenix, I would need to know more about what parameters you are calculating and your data. Prior tests comparing PKNCA to Phoenix have found almost no differences, and where the differences exist, they are intentional (AUC for all 0 concentrations and an edge case of half lives around order of automatically selecting the best half life). |
Beta Was this translation helpful? Give feedback.
-
|
I have prepared a new file with made up values but that respects the structure I am using. I am not able to upload the file here for some reason - so I will leave here the results.
I performed a NCA analysis with phoenix and these are the results I am expecting: NCA IV: (ex mouse 62) 0.993 ; 0.132 ; 1070 ; 1070 ; 1740 ; 1.00 ; 347; 347 ; 349 ; 349 ; 523 ; 2870 NCA PO: (ex mouse 67) - ; - ; 1.00; 8.46 ; 0.846 ; 1.48 ; 1.72 ; 17.2 ; - ; - ; - Thank you so much for your valuable feedback! |
Beta Was this translation helpful? Give feedback.
-
|
I'm confused about your results. When looking at the results for mouse 62, the Cmax_D value from Phoenix is 735, but the observed Cmax at time 0.083 is 1070 and the dose is 1 mg/kg. And, Clast is 14.6 from Phoenix but the observed Clast is 8.16. Given the mismatch between the data and the simple to extract results, I'm not sure how I can help. |
Beta Was this translation helpful? Give feedback.
-
|
you are completely right! I copied the wrong results. I am sorry for my mistake. My question now doesn't lie so much on the results per si, as with this structure of data I am not even able to get results with PKNCA. Plus, I wanted to understand if I should continue to use sparse as I don't think this structure requires. Thank you so much for your guidance! |
Beta Was this translation helpful? Give feedback.
-
|
My strategy now lies around the fact that I am extracting both concentration and dose info from the same set of data and the time associated to both is extracted from the same column. In your examples you use d_conc whith a different set of time date from d_dose, and I was wondering it that could be the problem to my lack of results, as if I am reading the data wrongly.
Thank you for your patience. |
Beta Was this translation helpful? Give feedback.
-
|
the code I wrote: # define tools version + version check Check_Rv <- R.Version() if (Rversion == Check_Rv) { } else { ------ Environment preparation ----------------------------------------------------------------------- --Set the working directory --Define the R script file path --Specify the output file path --Check working directory --Load dataset --Visualise the first five rows of my data ---------- Data Preparation --------------------------------------------------------------------------- --Assuming 'IV' for intravascular and 'PO' for extravascular Routes --Convert values to numeric for both datasets oral_data$Concentration <- as.numeric(oral_data$Concentration) -- Replace NA values resulting from BLQ with zero --Filter dose data to include only rows where DoseTime is equal to 0 --Remove rows with missing values --Convert Dose and DoseTime to numeric o_conc <- PKNCAconc(data=iv_data, formula=Concentration~Time|Subject) --Put your filtered dose data into a PKNCAdose object --Combine the two (and automatically determine the intervals of interest) --Compute the NCA parameters --Summarize the results versus the results: Caption: auclast, cmax, aucinf.obs: geometric mean and geometric coefficient of variation; tmax: median and range; half.life: arithmetic mean and standard deviation Warning messages: I don't see how to make this package work. Thank you. |
Beta Was this translation helpful? Give feedback.
-
|
https://cran.r-project.org/web/packages/PKNCA/vignettes/v04-sparse.html |
Beta Was this translation helpful? Give feedback.
-
|
That's a good question, if your data are sparse, you must indicate that to PKNCA. |
Beta Was this translation helpful? Give feedback.
-
|
I am not, I thought this was not an example when I had to use sparse, maybe I understood it wrong. I have in fact 3 mice that are administered only one dose and during a 24h interval suffer 5 plasma sample collection each to measure the concentration over this time slot. I tried to change it to: iv_conc <- PKNCAconc(iv_data, Concentration~Time|Subject+Dose, sparse=TRUE) iv_dose <- PKNCAdose(iv_data, Dose~Time|Subject+Dose, "intravascular", sparse=TRUE) d_interval <- data.frame(start = 0.083, iv_data <- PKNCAdata(iv_conc,iv_dose, intervals=d_interval) o_result <- pk.nca(iv_data) But I still get this Error: Error in standardize_column_names(x = data_no_exclude, cols = cols, group_cols = group_cols, : |
Beta Was this translation helpful? Give feedback.
-
|
Sorry for the slow reply on this. You cannot use a column name for both a grouping variable and one of the variables like time, concentration, or dose. The reason for this is that under the surface, PKNCA uses the |
Beta Was this translation helpful? Give feedback.
-
|
If you can think of a clearer way to express that error message, please let me know. |
Beta Was this translation helpful? Give feedback.
The first issue is the main one. Without fixing it, I'm not sure how you got results of any type.
The way that you specified the call to PKNCAconc, you didn't indicate a subject column. Given your description of the study, that is the most likely issue. The way to specify the subject column is to add the subject as the last grouping level similar to the.below call:
To get the group mean, you would run
summary()on your NCA results object.For the difference with Phoenix, I would need to know more about what parameters you are calculating and your data. Prior tests comparing PKNCA to Phoenix have found almost no differences, a…