aNCA Manual

Overview

This manual is a user guide and technical reference on how aNCA operates internally. It covers data requirements, workflow, settings and output definitions.

For a quick start and installation instructions, see Get Started. For NCA concepts, see What is NCA?.

The NCA workflow consists of five main steps executed through the aNCA interface.

Upload & Map Data: Load your dataset and map columns to ADNCA format using the Data tab
Configure Settings: Choose imputation methods, BLQ handling, and AUC calculation in the NCA Setup tab
Review Intervals: Verify dose times and intervals are correctly identified
Run Analysis: Click “Run NCA” to execute PKNCA calculations
Generate Outputs: See all results in the NCA results tab and export them to a ZIP using the top-right button.

Main Functions

The workflow is orchestrated internally by:

Data Preparation: apply_mapping(), format_pkncaconc_data(), format_pkncadose_data()
Interval Creation: format_pkncadata_intervals(), update_main_intervals()
NCA Execution: PKNCA’s PKNCAdata() and pk.nca()
Output Generation: export_cdisc() for CDISC datasets, TLG generation functions

1. Data Requirements

1.1 Required Columns

Variable	Label	Description	CDISC Reference
`STUDYID` *	Study Identifier	Character column to identify the study	CDISC Standard
`USUBJID` *	Unique Subject Identifier	Character/numeric column to identify a subject across studies	CDISC Standard
`ATPTREF` *	Analysis Timepoint Reference	Can be any column, used to filter data for NCA calculations (e.g., visit, dose number)	CDISC ADNCA

The sample identifiers help define what records are used for NCA calculations. At minimum, you need a parameter (PARAM), specimen type (PCSPEC), and concentration value (AVAL). - The specimen type (PCSPEC) is used to differentiate between different sample matrices (e.g., plasma vs urine) and allows for separate NCA excretion calculations for urine if needed. - The parameter (PARAM) is used to identify the analyte being measured and allows for multi-analyte assays to be processed in the same dataset. - Both generally together also help specifying the correct unit (AVALU) for the concentration values (AVAL).

Variable	Label	Description	CDISC Reference
`PARAM` *	Parameter	Name of the measured analyte	CDISC Standard
`PCSPEC` *	Specimen Material Type	Sample matrix (PLASMA, SERUM, URINE, etc.)	CDISC Standard
`AVAL` *	Analysis Value	Numeric analyte concentration (BLQ values assumed to be 0 if not imputed)	CDISC Standard

1.1.1 Additional Optional Variables

AEFRLT and AFRLT are useful for the PK excretion parameters that need to consider the duration of sample collection to calculate rates or total amounts.
ADOSEDUR is used to determine the appropriate start imputation for missing C₀ values in intravascular doses (IV). The automatic half-life determined by PKNCA may also change based on the duration of the infusion for superior accuracy. If not provided, aNCA assumes all doses are bolus (ADOSEDUR = 0) and applies the corresponding imputation rules.
VOLUME is used for excretion studies to calculate the total amount excreted in urine or feces. It can also be used to calculate the excretion rate if the sample collection duration is provided (via AEFRLT).
WTBL is used for excretion studies to calculate dose-normalized parameters based on body weight.
TRTRINT is commonly known as Tau (τ) and is used for the last dose of a subject/treatment (e.g., start + TRTRINT). If not provided, aNCA will guess the dosing interval based on the exact actual dose times, applying uneven intervals if necessary.
METABFL is used to flag records that represent metabolites. This allows users to easily identify and analyze metabolite data separately from parent drug data to compute metabolite-ratio operations. Additionally, it will also be considered for the start imputation rules, where metabolites will be imputed to C0 = 0 for first doses regardless of the route of administration or the availability of dose duration information.

Variable	Label	Description	Use Case
`AEFRLT`	Actual Relative End Time from First Dose	End time of sample collection	Urine/feces samples
`ADOSEDUR`	Actual Duration of Treatment Dose	Dose infusion duration	IV infusions
`VOLUME`	Volume Value	Collected sample volume	Excretion studies
`VOLUMEU`	Volume Value Unit	Unit of VOLUME	Excretion studies
`WTBL`	Weight at Baseline	Subject’s weight	Excretion studies
`WTBLU`	Weight at Baseline Unit	Unit of WTBL	Excretion studies
`TRTRINT`	Planned Treatment Interval	Time between doses (τ) in multiple dose studies	Multiple dosing
`METABFL`	Metabolite Flag	“Y” if PARAM is a metabolite	Multi-analyte assays

1.2 Data Format Example

The dose variables describe the treatment administration information necessary for NCA calculations.

DOSETRT helps us identify different treatments even if they happen within the same subject.
ROUTE is used to derive the PKNCA route variable (intravascular, extravascular). This, together with the dose duration (ADOSEDUR), helps determine the appropriate start imputations for missing C₀ values.
DOSEA is used for dose normalization of parameters like Cmax and AUC if the user selects that option or the ...DN parameters, apart from other parameters that are always dose-normalized (e.g., clearance).

Variable	Label	Description	CDISC Reference
`DOSETRT` *	Name of Treatment	Protocol-specified study treatment	CDISC Standard
`ROUTE` *	Route of Administration	Either “intravascular” or “extravascular” (or specific routes)	C66729
`DOSEA` *	Actual Treatment Dose	Actual dose amount given	CDISC Standard

2. Data Preparation

Data preparation involves mapping your raw dataset to the ADNCA format, handling column naming and units, and setting up exclusion rules before NCA analysis.

2.1 Column Mapping

Column mapping translates your raw dataset into ADNCA format. This is done via the UI mapping interface or the apply_mapping() function.

Key Operations: - Rename columns to match ADNCA standards (e.g., “CONCENTRATION” → AVAL) - Create default values for unmapped required variables (e.g., ADOSEDUR = 0 if not provided) - Apply CDISC variable labels and formatting - Remove duplicate concentration records - Handle special cases: - If DOSETRT not mapped, sets DOSETRT = PARAM - If ADOSEDUR not mapped, sets all values to 0

2.2 Time Variables

Time variables relate dose administrations to sample collection times. These are essential for proper interval calculations in NCA.

Variable	Label	Description
`AFRLT` *	Actual Relative Time from Analyte First Dose	Actual time since first dose
`ARRLT` *	Actual Relative Time from Reference Dose	Actual time since last dose
`NFRLT` *	Nominal Relative Time from Analyte First Dose	Planned time since first dose
`NRRLT` *	Nominal Relative Time from Reference Dose	Planned time since last dose

2.3 Units

Units are essential for parameter calculations and reporting. All unit columns must be present and consistent.

Variable	Label	Description
`AVALU` *	Analysis Value Unit	Unit of AVAL (e.g., “ng/mL”, “µg/mL”)
`RRLTU` *	Relative Time Unit	Unit of time variables (e.g., “hr”, “day”)
`DOSEU` *	Dose Units	Unit of DOSEA (e.g., “mg”, “mg/kg”)

2.4 Record Exclusions (NCAwXRS)

The NCAwXRS columns are used to exclude specific records from NCA calculations at the data preparation stage.

Column Name Pattern: NCA1XRS, NCA2XRS, …, NCA30XRS

Purpose: Character columns that explain why a record should not be included in PK NCA calculations.

Effects on NCA Computation: - Records with any non-empty value in NCAwXRS columns are automatically excluded from all PK calculations - Exclusion reasons are concatenated and stored in the internal nca_exclude column - Excluded records are retained in outputs but marked as excluded - Multiple exclusion columns allow different reasons to be tracked separately

Example:

ADNCA$NCA1XRS <- ifelse(ADNCA$AVAL > 10000, "Concentration too high", "")
ADNCA$NCA2XRS <- ifelse(ADNCA$ARRLT < 0, "Negative time point", "")

2.5 Data Preparation Implementation

Data preparation is orchestrated through aNCA’s internal functions:

2.5.1 Column Mapping: `apply_mapping()`

Purpose: Transforms user dataset to ADNCA format

Operations: - Renames columns based on user-specified mappings - Creates new columns for unmapped required variables (e.g., ADOSEDUR = 0 if not mapped) - If DOSETRT not mapped, sets DOSETRT = PARAM; if ADOSEDUR not mapped, sets all values to 0 - Removes concentration duplicates - Applies CDISC variable labels

2.5.2 Concentration & Dose Data Formatting

format_pkncaconc_data(): Creates PKNCA-compatible concentration dataset - Filters out dosing records - Creates DOSNOA: Sequential dose number within each group - Standardizes route (“intravascular” vs “extravascular”) - Merges NCAwXRS exclusion reasons into nca_exclude column

format_pkncadose_data(): Extracts and formats dosing information - Creates dose records from concentration data - Calculates sequential dose numbers - Links dose times to concentration profiles - Dose times are calculated based on AFRLT - ARRLT rounding to the 6th decimal to prevent floating point issues in interval creation.

2.5.3 Interval Creation & Parameter Assignment

format_pkncadata_intervals() & update_main_intervals(): Defines NCA intervals - Creates intervals based on dose times (single dose: [0, Inf]; multiple dose: [dose_time, next_dose_time or start + TRTRINT]) - Assigns PKNCA parameters based on study type (single/multiple dose, IV/EV) - Implements BLQ and start concentration imputation rules

3. NCA Settings

NCA settings control how PKNCA processes your data during calculations. These are configured in the NCA Setup tab.

3.1 Data Imputation Rules

The imputation actions are specified exclusively for parameters that are not exclusively calculated based on observed concentrations (e.g., Cmax, Tmax, AUClast, etc.) but also for parameters that require extrapolation to infinity (e.g., AUCinf) or parameters that are calculated based on the terminal slope (e.g., half-life, clearance, volume). The imputation rules are applied during the interval creation step in update_main_intervals(), where the appropriate imputation method is selected based on the characteristics of each record (e.g., dose route, dose duration, metabolite status, etc.) and the availability of data (e.g., presence of pre-dose concentrations, presence of multiple post-dose concentrations for log-linear regression). They are specified in the PKNCAdata object via the impute column. The order of imputation is as follows:

Order of Application: start_concentration -> BLQ imputation

3.1.1 Start Concentration Imputation: `create_start_impute()`

File: R/create_start_impute.R

Purpose: Defines how to handle missing or implied start concentrations (C₀)

Imputation rules specify how to handle missing or implied start concentrations (C₀). Selection depends on: - Dose route (IV vs extravascular) - Dose duration (bolus vs infusion) - Analyte type (parent drug vs metabolite) - Data availability (e.g., concentration at dose available, pre-dose concentrations close to dose…)

Condition	Imputation Method	Description
C₀ measured	`NA`	No imputation needed
1st dose, not IV bolus	`start_conc0`	C₀ = 0
1st dose, metabolite	`start_conc0`	C₀ = 0
Later dose, not IV bolus	`start_predose`	C₀ = pre-dose concentration
Later dose, metabolite	`start_predose`	C₀ = pre-dose concentration
IV bolus, parent drug, logslope possible	`start_logslope`	Back-extrapolate using log-linear regression
IV bolus, parent drug, logslope not possible	`start_c1`	Use first measured concentration

PKNCA Method: pk.calc.c0() from PKNCA package

3.1.2 BLQ (Below Limit of Quantification) Imputation

Implementation: Via blq_imputation_rule parameter in update_main_intervals()

Supported Strategies:

Position-based (PKNCA default):
- first: How to handle BLQ at start of profile
- middle: How to handle BLQ in middle of profile
- last: How to handle BLQ at end of profile
Tmax-based:
- before.tmax: BLQ before maximum concentration
- after.tmax: BLQ after maximum concentration

Options for each position:

Numeric value (e.g., 0, LOQ/2): Substitute BLQ with this value
"drop": Exclude the BLQ value from calculations
"keep": Use BLQ as 0

Example:

blq_rule <- list(
  first = "0",        # Set first BLQ to 0
  middle = "drop",    # Drop middle BLQ values
  last = "keep"       # Keep last BLQ as 0
)

3.2 BLQ (Below Limit of Quantification) Handling

BLQ values are handled using position-based or Tmax-based strategies.

Position-based Options: - first: Strategy for BLQ at profile start - middle: Strategy for BLQ in profile middle - last: Strategy for BLQ at profile end

For each position, you can choose: - Numeric value (e.g., 0, LOQ/2): Substitute BLQ with this value - "drop": Exclude from calculations - "keep": Treat as 0

3.3 AUC Calculation Methods

aNCA supports multiple AUC calculation methods via PKNCA. These methods control how the area under the concentration-time curve is calculated.

3.3.1 Integration Methods (Between Concentrations)

aNCA supports three methods for interpolating AUC between two concentration measurements:

Method	Interpolation Rules	Best For
Linear-up/log-down (default)	Linear for rising, logarithmic for declining concentrations	Standard/exogenous substances with first-order elimination
Linear	Linear trapezoidal rule for all segments	Endogenous substances or non-first-order elimination
Lin-log	Linear before Tmax, logarithmic after Tmax	Infrequently used; specific modeling scenarios

Details: - Linear-up/log-down: Most commonly used in PK. Applies linear interpolation when concentrations increase or when either concentration is zero; uses log interpolation for declining concentrations. - Linear: Simplest method; uses linear trapezoidal rule throughout. Good for situations where first-order kinetics may not apply. - Lin-log: Uses linear interpolation before Tmax, logarithmic interpolation after Tmax.

3.3.2 Extrapolation Methods (After Last Measurement)

After the last measurable concentration (Tlast), AUC can be extrapolated using different approaches:

Method	Description	Use Case
AUClast	No extrapolation; extrapolated AUC = 0	When full profile to negligible concentrations is available
AUCall	Linear triangle from Tlast to zero	When Tlast is not the final measurement; interpolates missing data
AUCinf	Extrapolates to infinity using: Clast/λz	When terminal elimination is well-characterized; common for regulatory submissions

Important Note: - aucinf.obs uses observed Clast (last measured concentration) - aucinf.pred uses predicted Clast (extrapolated via regression) - High extrapolation percentage (>20%) may indicate poor terminal phase characterization

3.3.3 AUC Types Available

Parameter	Description	Extrapolation	Formula
`auclast`	AUC to last measurable concentration	None	Sum of trapezoids to Tlast
`aucall`	AUC to infinity with linear extrapolation	Linear triangle	AUClast + triangle(Tlast→0)
`aucinf.obs`	AUC to infinity (observed Clast)	Using Clast directly	AUClast + Clast/λz
`aucinf.pred`	AUC to infinity (predicted Clast)	Using regression fit	AUClast + Ĉlast/λz
`aucint`	Partial AUC over specified interval	User-defined	AUC(time1, time2)

4. Half-Life Selection and Slope Selector

4.1 Automated Half-Life Calculation

By default, PKNCA automatically selects the terminal elimination phase using the following algorithm:

Omit all missing and BLQ concentrations. If the administration is an infusion, also omit all points during the infusion.
Estimate the half-life for each set of points from the first concentration measure after Tmax to the third measure before Tlast.
Select the best half-life expecting that:

The adjusted r-squared must be within a tolerance factor (typically 0.0001) of the largest adjusted r-squared.
The λz value (slope for the half-life line) must be positive; in other words, the half-life slope must be decreasing.
If multiple choices of points fit the above criteria, choose the one with the most concentration measurements included.

Reference: PKNCA Half-Life Calculation

4.2 Manual Slope Selection

Users can override automatic selection using the Slope Selector tab in the NCA Setup.

4.2.1 Inclusion (Manual Slope Selection)

Purpose: Specify exactly which points to use for half-life calculation

Method:

Via UI: Click first and last points to include
Via Table: Enter subject, analyte, and RANGE (point indices)

Effect:

Overrides PKNCA’s automatic selection
Uses only the specified points for half-life regression
Requires at least 3 points

Implementation:

Sets is.included.hl = TRUE for selected points
Uses PKNCA’s include_half.life column
Function: PKNCA uses PKNCAconc(..., include_half.life="include_half.life")

4.2.2 Exclusion (Point Removal)

Purpose: Remove outlier points from half-life calculation

Method:

Via UI: Double-click points to exclude
Via Table: Enter subject, analyte, and RANGE

Effect:

Removes specified points from half-life calculation
PKNCA recalculates using remaining points
Can exclude multiple points

Implementation:

Sets is.excluded.hl = TRUE for excluded points
Uses PKNCA’s exclude_half.life column
Function: PKNCA uses PKNCAconc(..., exclude_half.life="exclude_half.life")

4.2.3 Viewing Selections

Results are shown in:

Slopes Information tab: Shows automatic and manual slope selections
Manual Adjustments tab: Lists user-made changes
Slope plots: Visual display of selected points

5. Running NCA

aNCA builds on PKNCA for core NCA calculations. Understanding PKNCA methods helps interpret results.

5.1 Data Object Creation

Function	Purpose	aNCA Wrapper
`PKNCA::PKNCAconc()`	Creates concentration object	`PKNCA_create_data_object()`
`PKNCA::PKNCAdose()`	Creates dose object	`PKNCA_create_data_object()`
`PKNCA::PKNCAdata()`	Combines conc + dose + intervals	`PKNCA_create_data_object()`

5.2 NCA Calculations

Function	Parameter	aNCA Usage
`pk.calc.cmax()`	Maximum concentration	Automatic
`pk.calc.tmax()`	Time of maximum	Automatic
`pk.calc.auclast()`	AUC to last	Automatic
`pk.calc.aucinf.obs()`	AUC to infinity	Automatic
`pk.calc.half.life()`	Terminal half-life	Automatic + manual
`pk.calc.cl.obs()`	Clearance	Automatic
`pk.calc.vz.obs()`	Volume of distribution	Automatic

Reference: PKNCA Function Reference

5.3 NCA Execution

# Simplified workflow
pknca_data <- PKNCAdata(conc, dose, intervals)
results <- pk.nca(pknca_data)

aNCA Wrapper: The NCA workflow is encapsulated in the app’s “Run NCA” button, which internally calls `pk.nca()` with configured settings.

6. Results & Outputs

All output files can be downloaded after running NCA. The export button is located in the top-right as Save.

6.1 Output Structure

The output package includes:

Exploration Outputs

Individual plots: Subject-level concentration-time profiles
Mean plots: Mean ± SD/SE by treatment group

NCA Results

nca_results: Pivoted PKNCA results for all calculated parameters with units

CDISC Datasets

PP: Pharmacokinetic Parameters (standard CDISC/SDTMIG format)
ADPP: Analysis Dataset for PK Parameters (ADaM format)
ADNCA: Input data (adapted to user selections with standardizations)

Additional Analysis

Matrix/specimen ratios: Metabolite/parent ratios or cross-matrix comparisons
Excretion results: Urine/feces data if applicable

Extras

R script: Reproduces entire analysis starting from raw data
Settings file: Defines analysis specifications (can be re-uploaded to re-run)
Results slides: Presentation deck with Exploration outputs and parameter summaries

6.2 Output Formats

Graphics and plots: PNG or HTML
Data tables: CSV, XPT, or RDS
Slide decks: PPTX or QMD (renderable to HTML/PDF)

6.3 CDISC Dataset Derivation

aNCA generates three CDISC-compliant output datasets:

PP (Pharmacokinetic Parameters): - Standard dataset containing calculated NCA parameters - Derivation: Extracts NCA results from PKNCA, translates to CDISC terms (PPTESTCD, PPTEST), reshapes to one row per subject-analyte-interval - Key columns: PPTESTCD (parameter code), PPTEST (parameter name), PPSTRESN (numeric result), PPSTRESU (unit), PPSPEC (specimen type)

ADPP (Analysis Dataset for Pharmacokinetic Parameters): - Analysis-ready version of PP with additional analysis variables - Derivation: Adds analysis flags (ANLFL, ANL01FL), grouping variables, treatment (TRT01A, TRT01P), and demographic variables (AGE, SEX, RACE) - Key addition: AVAL (analysis value, copy of PPSTRESN)

ADNCA (Input Dataset): - Input dataset maintained in output for traceability - Minimal modifications: standardized route, dose numbers assigned - Excluded records retained but marked - Purpose: Traceability and support for concentration-time plotting and re-analysis

7. TLG (Tables, Listings, and Graphs) Framework

Users can customize TLGs using widgets in the TLG tab:

Titles and Footnotes: Free text input
Axis Labels: Customizable with special syntax:
- !COLUMN references column label
- $COLUMN references column value
Grouping Variables: Select from available columns
Statistical Options: Choose statistics to display

Example:

X-axis label: Time (!RRLTU)
# Expands to: Time (hr) if RRLTU = "hr"
Title: Study $STUDYID - !PARAM Analysis
# Expands to: Study STUDY001 - Drug X Analysis

7.1 TLG Generation Process

Order Details: User selects desired TLGs
Parameter Selection: For each TLG, specify options
Generation: TLGs created using appropriate functions
Export: Download as PDF, Word, or PowerPoint

Implementation:

TLG templates defined in inst/shiny/tlg_definitions/
Generation functions in R/ (e.g., g_pkcg.R, l_pkcl01.R)
Export handled by officer and flextable

8. Troubleshooting & FAQ

8.1 Common Issues

Issue: “No intervals for data” error

Cause: No valid NCA intervals were created from the dose and concentration data.

Solutions: - Verify dose times are present in the data - Check that time variables (AFRLT, ARRLT) are properly mapped and numeric - Ensure at least some concentrations fall within expected intervals - Review dose-timepoint relationships for logical errors

Issue: “Too few points for half-life”

Cause: Insufficient concentration measurements to calculate terminal slope.

Solutions: - Increase number of concentration samples post-Tmax - Ensure adequate sampling in terminal phase (final 25% of profile) - Check BLQ handling - missing early concentrations may affect calculations - Consider manual slope selection if automatic selection fails

Issue: Half-life dependent parameters flagged

Cause: Terminal slope calculation doesn’t meet quality criteria (r² < 0.9, span ratio < 2, etc.).

Solutions: - Use manual slope selection to override automatic selection - Review raw data for outliers or data quality issues - Consider alternative AUC methods that don’t depend on half-life extrapolation - Check if BLQ handling is appropriate

Issue: High AUC extrapolation percentage

Cause: Large portion of AUC from extrapolation beyond measured data.

Solutions: - Extend follow-up sampling period if possible - Use observed Clast method instead of predicted - Review whether extrapolation assumption is valid - Consider flagging parameters if extrapolation exceeds acceptable thresholds

8.2 Understanding Result Flags

aNCA automatically flags results based on predefined scientific rules to help identify potential data or calculation issues.

Half-Life Selection Flags (indicates terminal slope quality): - r.squared (R2) < 0.9: Poor fit of terminal slope - adj.r.squared < 0.8: Poor adjusted fit of terminal slope - span.ratio < 2: Insufficient span of points for reliable estimation

AUC Extrapolation Flags (indicates extrapolation beyond measured data): - auc_extrap_pct_obs > 20%: High extrapolation based on observed Clast - auc_extrap_pct_pred > 20%: High extrapolation based on predicted Clast

Flagged parameters should be reviewed to determine if they’re reliable for decision-making. Parameters dependent on flagged half-lives (e.g., AUCinf, clearance) will inherit those flags.

8.3 Where to Get Help

GitHub Issues: Report bugs or request features at https://github.com/pharmaverse/aNCA/issues
Documentation: See the Get Started guide and Background vignette
Community: Join the Pharmaverse community for discussions

9. References & Resources

9.1 PKNCA Documentation

Main Site: https://billdenney.github.io/pknca/
Half-Life Selection: https://billdenney.github.io/pknca/articles/Half-life.html
AUC Calculations: https://billdenney.github.io/pknca/reference/pk.calc.auc.html
Parameter Reference: https://billdenney.github.io/pknca/reference/index.html#pk-parameters

9.2 CDISC Standards

SDTMIG: Study Data Tabulation Model Implementation Guide
ADaMIG-ADNCA: Analysis Data Model Implementation Guide
ADNCA: https://www.cdisc.org/standards/foundational/adam

9.3 Further Reading

Denney, W. S., Duvvuri, S., & Buckeridge, C. (2015). Simple, Automatic Noncompartmental Analysis: The PKNCA R Package. Journal of Pharmacokinetics and Pharmacodynamics, 42(1), 11-17. https://doi.org/10.1007/s10928-015-9432-2

10. Glossary

Term	Definition
ADNCA	Analysis Dataset for Non-Compartmental Analysis (CDISC format)
ADPP	Analysis Dataset for Pharmacokinetic Parameters
AUC	Area Under the Curve - measure of drug exposure
BLQ	Below Limit of Quantification
CDISC	Clinical Data Interchange Standards Consortium
Cmax	Maximum observed concentration
Half-life (t½)	Time for concentration to decrease by 50%
NCA	Non-Compartmental Analysis
PKNCA	R package for pharmacokinetic non-compartmental analysis
PP	Pharmacokinetic Parameters domain
TLG	Tables, Listings, and Graphs
Tmax	Time of maximum concentration
λz	Terminal elimination rate constant

For questions or issues, please visit the GitHub repository or consult the FAQ page.

Overview

Main Functions

1. Data Requirements

1.1 Required Columns

1.1.1 Additional Optional Variables

1.2 Data Format Example

2. Data Preparation

2.1 Column Mapping

2.2 Time Variables

2.3 Units

2.4 Record Exclusions (NCAwXRS)

2.5 Data Preparation Implementation

2.5.1 Column Mapping: apply_mapping()

2.5.2 Concentration & Dose Data Formatting

2.5.3 Interval Creation & Parameter Assignment

3. NCA Settings

3.1 Data Imputation Rules

3.1.1 Start Concentration Imputation: create_start_impute()

3.1.2 BLQ (Below Limit of Quantification) Imputation

3.2 BLQ (Below Limit of Quantification) Handling

3.3 AUC Calculation Methods

3.3.1 Integration Methods (Between Concentrations)

3.3.2 Extrapolation Methods (After Last Measurement)

3.3.3 AUC Types Available

4. Half-Life Selection and Slope Selector

4.1 Automated Half-Life Calculation

4.2 Manual Slope Selection

4.2.1 Inclusion (Manual Slope Selection)

4.2.2 Exclusion (Point Removal)

4.2.3 Viewing Selections

5. Running NCA

5.1 Data Object Creation

5.2 NCA Calculations

5.3 NCA Execution

aNCA Wrapper: The NCA workflow is encapsulated in the app’s “Run NCA” button, which internally calls pk.nca() with configured settings.

6. Results & Outputs

6.1 Output Structure

Exploration Outputs

NCA Results

CDISC Datasets

Additional Analysis

Extras

6.2 Output Formats

6.3 CDISC Dataset Derivation

7. TLG (Tables, Listings, and Graphs) Framework

7.1 TLG Generation Process

8. Troubleshooting & FAQ

8.1 Common Issues

Issue: “No intervals for data” error

Issue: “Too few points for half-life”

Issue: Half-life dependent parameters flagged

Issue: High AUC extrapolation percentage

8.2 Understanding Result Flags

8.3 Where to Get Help

9. References & Resources

9.1 PKNCA Documentation

9.2 CDISC Standards

9.3 Further Reading

10. Glossary

2.5.1 Column Mapping: `apply_mapping()`

3.1.1 Start Concentration Imputation: `create_start_impute()`

aNCA Wrapper: The NCA workflow is encapsulated in the app’s “Run NCA” button, which internally calls `pk.nca()` with configured settings.