Package 'admetshiny' reference manual

Title:	Interactive ADMET and Drug-Likeness Analysis of Small Molecules
Description:	Provides an interactive Shiny application and a toolbox of R functions for the management, calculation, filtering, visualization and exploratory analysis of molecular descriptors and ADMET (Absorption, Distribution, Metabolism, Excretion and Toxicity) properties of small molecules. Computes descriptors locally via the Chemistry Development Kit (CDK), and offers drug-likeness filters (Lipinski, Veber, Ghose, Egan, Muegge), the BOILED-Egg model for gastrointestinal absorption and blood-brain barrier permeability, a P-glycoprotein (P-gp, also known as ATP-binding cassette sub-family B member 1, ABCB1) substrate Random Forest classifier, Principal Component Analysis (PCA), t-Distributed Stochastic Neighbor Embedding (t-SNE), Uniform Manifold Approximation and Projection (UMAP), radar plots and Tanimoto / AGglomerative NESting (AGNES) clustering to support compound prioritization in early-stage drug discovery.
Authors:	Xavier Clemente Garcia Cevallos [aut, cre] (ORCID: <https://orcid.org/0000-0002-3284-2363>)
Maintainer:	Xavier Clemente Garcia Cevallos <[email protected]>
License:	MIT + file LICENSE
Version:	1.0.0
Built:	2026-10-01 11:08:01 UTC
Source:	https://github.com/cran/admetshiny

admetshiny: Interactive ADMET and Drug-Likeness Analysis

Description

The admetshiny package provides an interactive Shiny application and a collection of R functions for the management, calculation, filtering, visualization and exploratory analysis of molecular descriptors and ADMET properties of small molecules. It computes descriptors locally via the Chemistry Development Kit (CDK) through rcdk, and offers drug-likeness filters (Lipinski, Veber, Ghose, Egan, Muegge), the BOILED-Egg model, a P-glycoprotein substrate Random Forest classifier, PCA, t-SNE, UMAP, parallel coordinates, radar plots and Tanimoto/AGNES structural similarity clustering.

Details

To launch the interactive application simply run:

admetshiny::run_app()

The exported functions (filters, plots, data normalisation and CDK helpers) can also be used programmatically in plain R scripts. By Xavier Clemente Garcia Cevallos :3 "Hecho con amor para la ciencia"

Drug-likeness filters

lipinskiFilter - Lipinski Rule-of-Five
veberFilter - Veber oral bioavailability
ghoseFilter - Ghose qualifying range
eganFilter - Egan PSA-LogP rule
mueggeFilter - Muegge pharmacophore filter
applyFilters - Apply multiple filters in sequence
computeViolationColumns - Compute violation columns

Data normalization

mapADMETColumns - Map any user dataset to standard schema
mapCDKDescriptors - Map CDK descriptors to standard schema

CDK & webchem

getSmilesFromIdentifiers - Retrieve SMILES from PubChem
calcCDKDescriptors - Calculate descriptors with CDK
computeADMETProperties - Compute BOILED-Egg ADMET properties

Visualizations

plotBoiledEgg - BOILED-Egg model
plotMW, plotTPSA, plotLogP - Distributions
plotPCA - PCA chemical space
plotTSNE - t-SNE chemical space
plotUMAP - UMAP chemical space
plotParallel - Parallel coordinates
plotViolin - Violin plot
plotRadar - Radar chart
plotTanimoto - Tanimoto/AGNES clustering
plotCorrHeatmap - Correlation heatmap
plotClusterHeatmap - Cluster heatmap with dendrogram
plotHistogramCustom - Customizable histogram of any numeric column
palette_selector_ui, apply_palette - Colour customization

Application

run_app - Launch the Shiny application
generateReport - Generate a report from the R console

Author(s)

Maintainer: Xavier Clemente Garcia Cevallos [email protected] (ORCID)

Authors:

Xavier Clemente Garcia Cevallos [email protected] (ORCID)

Apply a colour palette to a ggplot object

Description

Modifies an existing ggplot object by adding a colour/fill scale matching the selected palette, intelligently detecting whether the colour variable is continuous or discrete.

Usage

apply_palette(p, palette_name, data, color_col)
apply_palette(p, palette_name, data, color_col)

Arguments

p

A ggplot object.

palette_name

Character. Name of the palette (as returned by palette_selector_ui).

data

The data.frame used to build p.

color_col

Character. Name of the column used for colouring, or "None".

Value

A ggplot object (possibly modified).

Examples

library(ggplot2)
d <- data.frame(MW = c(300, 400, 500), group = c("A", "A", "B"))
p <- ggplot(d, aes(x = group, y = MW, color = group)) + geom_point()
p <- apply_palette(p, "Set1", d, "group")
library(ggplot2)
d <- data.frame(MW = c(300, 400, 500), group = c("A", "A", "B"))
p <- ggplot(d, aes(x = group, y = MW, color = group)) + geom_point()
p <- apply_palette(p, "Set1", d, "group")

Apply drug-likeness filters in sequence

Description

Applies the selected drug-likeness filters to a normalized data.frame. The filters are applied in the order Lipinski, Veber, Ghose, Egan, Muegge; only those named in filters are executed.

Usage

applyFilters(
  data,
  filters,
  lipinski = list(),
  veber = list(),
  ghose = list(),
  egan = list(),
  muegge = list()
)
applyFilters(
  data,
  filters,
  lipinski = list(),
  veber = list(),
  ghose = list(),
  egan = list(),
  muegge = list()
)

Arguments

data

A data.frame already normalized (e.g. by mapADMETColumns, mapCDKDescriptors, or mapCDKDescriptors).

filters

Character vector with the names of the filters to apply. Any subset of c("Lipinski", "Veber", "Ghose", "Egan", "Muegge").

lipinski, veber, ghose, egan, muegge

Named lists of parameters forwarded to the corresponding filter function.

Value

A data.frame with the rows of data that pass every selected filter.

Examples

d <- data.frame(MW = 300, LogP = 2, TPSA = 80, MR = 90,
  "#Heavy atoms" = 22, "#Rotatable bonds" = 3,
  "#H-bond acceptors" = 4, "#H-bond donors" = 2,
  "Lipinski #violations" = 0, "Ghose #violations" = 0,
  "Veber #violations" = 0, "Egan #violations" = 0,
  "Muegge #violations" = 0, check.names = FALSE)
applyFilters(d, filters = c("Lipinski", "Veber", "Egan"))
d <- data.frame(MW = 300, LogP = 2, TPSA = 80, MR = 90,
  "#Heavy atoms" = 22, "#Rotatable bonds" = 3,
  "#H-bond acceptors" = 4, "#H-bond donors" = 2,
  "Lipinski #violations" = 0, "Ghose #violations" = 0,
  "Veber #violations" = 0, "Egan #violations" = 0,
  "Muegge #violations" = 0, check.names = FALSE)
applyFilters(d, filters = c("Lipinski", "Veber", "Egan"))

Calculate molecular descriptors with CDK

Description

Parses SMILES strings and calculates the requested molecular descriptors locally using the Chemistry Development Kit (CDK) through the suggested package rcdk. Requires a working Java JDK (a JRE alone is not sufficient).

Usage

calcCDKDescriptors(
  smiles,
  which = c("mw", "alogp", "tpsa", "hbd", "hba", "rotb", "heavy", "aroma", "mr")
)
calcCDKDescriptors(
  smiles,
  which = c("mw", "alogp", "tpsa", "hbd", "hba", "rotb", "heavy", "aroma", "mr")
)

Arguments

smiles

Character vector of SMILES strings.

which

Character vector of descriptor short names. Any subset of c("mw", "alogp", "tpsa", "hbd", "hba", "rotb", "heavy", "aroma", "mr"). By default all are calculated.

Value

A data.frame with one row per valid molecule and a SMILES column.

Examples


smiles <- c("CCO", "CC(=O)Oc1ccccc1C(=O)O")
desc <- calcCDKDescriptors(smiles)

smiles <- c("CCO", "CC(=O)Oc1ccccc1C(=O)O")
desc <- calcCDKDescriptors(smiles)

Compute BOILED-Egg ADMET properties

Description

Computes the gastrointestinal absorption (GI absorption), blood-brain barrier permeability (BBB permeant) and P-glycoprotein substrate (Pgp substrate) categorical columns from the LogP and TPSA values, using the official BOILED-Egg polygon coordinates (Daina & Zoete, 2016, Data S3) for GI/BBB classification via point-in-polygon testing, and a Random Forest classifier for P-gp substrate prediction.

Usage

computeADMETProperties(data)
computeADMETProperties(data)

Arguments

data

A data.frame with LogP, TPSA, MW, #H-bond donors and #H-bond acceptors columns.

Details

The BOILED-Egg model was originally calibrated with WLOGP; here the application's generic LogP column is used (which may be WLOGP, Consensus Log P, ALogP, or the source platform's own LogP). This is an acceptable approximation for exploratory visualization.

Value

A data.frame with the added ADMET property columns.

References

Daina, A., & Zoete, V. (2016). A boiled egg to predict gastrointestinal absorption and brain penetration of small molecules. ChemMedChem, 11(11), 1117-1121.

Sedykh, A., Fourches, D., Duan, J., et al. (2013). Human intestinal transporter database: QSAR modeling and virtual excretion experiments. J. Cheminformatics 2015, 7:21 (Metrabase P-gp data).

Examples


d <- data.frame(LogP = 2, TPSA = 80, MW = 300,
  "#H-bond donors" = 2, "#H-bond acceptors" = 4, check.names = FALSE)
d <- computeADMETProperties(d)

d <- data.frame(LogP = 2, TPSA = 80, MW = 300,
  "#H-bond donors" = 2, "#H-bond acceptors" = 4, check.names = FALSE)
d <- computeADMETProperties(d)

Compute drug-likeness violation columns

Description

Computes the five drug-likeness "#violations" columns (Lipinski, Ghose, Veber, Egan, Muegge) from the physicochemical properties, using the thresholds from the original publications. Missing columns are safely filled with NA.

Usage

computeViolationColumns(data)
computeViolationColumns(data)

Arguments

data

A data.frame with physicochemical property columns.

Details

The generic LogP column is used for all LogP-dependent rules. The original publication thresholds are used:

Lipinski: LogP > 5 (original Rule-of-Five)
Ghose: LogP < -0.4 or LogP > 5.6
Egan: LogP > 5.88
Muegge: LogP < -2 or LogP > 5

For the Ghose atom-count criterion (20-70 atoms), the function uses "#Total atoms" (heavy + H, matching the original Ghose 1999 definition) when available, and falls back to "#Heavy atoms" for backwards compatibility.

The Veber violation is binary: a compound violates if it has more than 10 rotatable bonds OR if both polarity conditions fail (TPSA > 140 AND HBA + HBD > 12).

Value

A data.frame with the added violation columns.

Examples


d <- data.frame(MW = 300, LogP = 2, TPSA = 80, MR = 90,
  "#H-bond acceptors" = 4, "#H-bond donors" = 2,
  "#Rotatable bonds" = 3, "#Heavy atoms" = 22, check.names = FALSE)
d <- computeViolationColumns(d)

d <- data.frame(MW = 300, LogP = 2, TPSA = 80, MR = 90,
  "#H-bond acceptors" = 4, "#H-bond donors" = 2,
  "#Rotatable bonds" = 3, "#Heavy atoms" = 22, check.names = FALSE)
d <- computeViolationColumns(d)

Egan drug-likeness filter

Description

Filters compounds according to the Egan rule (Egan et al., 2000) based on TPSA and LogP.

Usage

eganFilter(data, e_tpsa = 131.6, e_logp = 5.88, violations = 0)
eganFilter(data, e_tpsa = 131.6, e_logp = 5.88, violations = 0)

Arguments

data

A data.frame with the standard column schema.

e_tpsa

Numeric. Maximum TPSA. Default 131.6.

e_logp

Numeric. Maximum LogP. Default 5.88.

violations

Integer. Maximum tolerated Egan violations. Default 0.

Value

A data.frame with the rows of data that satisfy all thresholds.

References

Egan, W. J., Merz, K. M., & Baldwin, J. J. (2000). Prediction of drug absorption using multivariate statistics. Journal of Medicinal Chemistry, 43(21), 3867-3877.

Examples

d <- data.frame(TPSA = 80, LogP = 2, "Egan #violations" = 0,
  check.names = FALSE)
eganFilter(d)
d <- data.frame(TPSA = 80, LogP = 2, "Egan #violations" = 0,
  check.names = FALSE)
eganFilter(d)

Generate an ADMETShiny report from the R console

Description

Generates a comprehensive ADMET and drug-likeness analysis report from one or more datasets, directly from the R console (without launching the Shiny app). The report includes per-dataset statistics, drug-likeness filter results, BOILED-Egg ADMET classification, additional literature-supported metrics, a composite drug-likeness score, cross-dataset comparison, visualizations and references.

Usage

generateReport(
  data,
  filters = character(0),
  format = "html",
  output_file = NULL,
  source_name = "Dataset"
)
generateReport(
  data,
  filters = character(0),
  format = "html",
  output_file = NULL,
  source_name = "Dataset"
)

Arguments

data

A data.frame, or a named list of data.frames. If a single data.frame is provided, it is treated as one dataset named "Dataset 1". If a list is provided, each element is a separate dataset and the list names are used as dataset labels.

filters

Character vector of drug-likeness filter names applied to the data (e.g. c("Lipinski", "Veber")). Use character(0) if no filters were applied. Default character(0).

format

Character; one of "html", "pdf", "doc". Default "html".

output_file

Character; path where the output file will be written. If NULL, a default name is used in the current working directory.

source_name

Character; human-readable name for the dataset. Only used when data is a single data.frame. Default "Dataset".

Value

Invisible NULL; called for the side-effect of writing the rendered report to output_file.

Examples


## Build a small synthetic ADMET dataset in the standard schema
d <- data.frame(
  Name = c("druglike", "violator"),
  SMILES = c("CCO", "CCCCCCCCCCCCCCCCCCCCCCCCCC"),
  MW = c(300, 650),
  LogP = c(2, 7),
  TPSA = c(40, 160),
  MR = c(70, 150),
  "#H-bond acceptors" = c(4, 12),
  "#H-bond donors" = c(2, 7),
  "#Rotatable bonds" = c(3, 14),
  "#Heavy atoms" = c(20, 80),
  "#Aromatic heavy atoms" = c(6, 9),
  check.names = FALSE
)
d <- computeViolationColumns(d)

## Generate an HTML report (written to a temp file)
out <- tempfile(fileext = ".html")
generateReport(d, filters = c("Lipinski", "Veber"),
               format = "html", output_file = out,
               source_name = "Example")

## Build a small synthetic ADMET dataset in the standard schema
d <- data.frame(
  Name = c("druglike", "violator"),
  SMILES = c("CCO", "CCCCCCCCCCCCCCCCCCCCCCCCCC"),
  MW = c(300, 650),
  LogP = c(2, 7),
  TPSA = c(40, 160),
  MR = c(70, 150),
  "#H-bond acceptors" = c(4, 12),
  "#H-bond donors" = c(2, 7),
  "#Rotatable bonds" = c(3, 14),
  "#Heavy atoms" = c(20, 80),
  "#Aromatic heavy atoms" = c(6, 9),
  check.names = FALSE
)
d <- computeViolationColumns(d)

## Generate an HTML report (written to a temp file)
out <- tempfile(fileext = ".html")
generateReport(d, filters = c("Lipinski", "Veber"),
               format = "html", output_file = out,
               source_name = "Example")

Retrieve canonical SMILES from PubChem

Description

Queries PubChem (via the suggested package webchem) to obtain CIDs and canonical SMILES for a vector of chemical identifiers (names, CAS numbers, InChIKeys or PubChem CIDs).

Usage

getSmilesFromIdentifiers(ids, from = "name")
getSmilesFromIdentifiers(ids, from = "name")

Arguments

ids

Character vector of identifiers.

from

Character. Type of identifier: one of "name", "cas", "inchikey", "cid". Default "name".

Value

A data.frame with columns query, cid, CanonicalSMILES, IsomericSMILES, MolecularFormula, IUPACName.

Examples

## Not run: 
# This service requires a constant internet connection and may fail if the
# server goes down. It may also involve long wait times. That's the main
# reason for using dontrun in this case.
ids <- c("aspirin", "ibuprofen")
smiles <- getSmilesFromIdentifiers(ids, from = "name")

## End(Not run)
## Not run: 
# This service requires a constant internet connection and may fail if the
# server goes down. It may also involve long wait times. That's the main
# reason for using dontrun in this case.
ids <- c("aspirin", "ibuprofen")
smiles <- getSmilesFromIdentifiers(ids, from = "name")

## End(Not run)

Ghose drug-likeness filter

Description

Filters compounds according to the Ghose qualifying range (Ghose et al., 1999): molecular weight, molar refractivity, LogP and number of atoms.

Usage

ghoseFilter(
  data,
  g_mw_min = 160,
  g_mw_max = 480,
  g_mr_min = 40,
  g_mr_max = 130,
  g_logp_min = -0.4,
  g_logp_max = 5.6,
  g_ha_min = 20,
  g_ha_max = 70,
  violations = 0
)
ghoseFilter(
  data,
  g_mw_min = 160,
  g_mw_max = 480,
  g_mr_min = 40,
  g_mr_max = 130,
  g_logp_min = -0.4,
  g_logp_max = 5.6,
  g_ha_min = 20,
  g_ha_max = 70,
  violations = 0
)

Arguments

data

A data.frame with the standard column schema.

g_mw_min

Numeric. Minimum molecular weight. Default 160.

g_mw_max

Numeric. Maximum molecular weight. Default 480.

g_mr_min

Numeric. Minimum molar refractivity. Default 40.

g_mr_max

Numeric. Maximum molar refractivity. Default 130.

g_logp_min

Numeric. Minimum LogP. Default -0.4.

g_logp_max

Numeric. Maximum LogP. Default 5.6.

g_ha_min

Numeric. Minimum number of atoms. Default 20.

g_ha_max

Numeric. Maximum number of atoms. Default 70.

violations

Integer. Maximum tolerated Ghose violations. Default 0.

Details

The "number of atoms" criterion (range 20-70) refers to TOTAL atoms (heavy + hydrogens), following the original Ghose (1999) definition.

Value

A data.frame with the rows of data that satisfy all thresholds.

References

Ghose, A. K., Viswanadhan, V. N., & Wendoloski, J. J. (1999). A knowledge-based approach in designing combinatorial or medicinal chemistry libraries for drug discovery. 1. A qualitative and quantitative characterization of known drug databases. Journal of Combinatorial Chemistry, 1(1), 55-68.

Examples

d <- data.frame(MW = 300, MR = 90, LogP = 2, "#Total atoms" = 30,
  "Ghose #violations" = 0, check.names = FALSE)
ghoseFilter(d)
d <- data.frame(MW = 300, MR = 90, LogP = 2, "#Total atoms" = 30,
  "Ghose #violations" = 0, check.names = FALSE)
ghoseFilter(d)

Lipinski Rule-of-Five filter

Description

Filters compounds according to the Rule of Five (Lipinski et al., 1997): molecular weight, LogP, number of H-bond acceptors and donors, plus the pre-computed number of Lipinski violations.

Usage

lipinskiFilter(data, mw = 500, logp = 5, hba = 10, hbd = 5, violations = 0)
lipinskiFilter(data, mw = 500, logp = 5, hba = 10, hbd = 5, violations = 0)

Arguments

data

A data.frame with the standard column schema (processed by mapADMETColumns, mapCDKDescriptors).

mw

Numeric. Maximum molecular weight. Default 500.

logp

Numeric. Maximum LogP. Default 5.

hba

Numeric. Maximum number of H-bond acceptors. Default 10.

hbd

Numeric. Maximum number of H-bond donors. Default 5.

violations

Integer. Maximum tolerated Lipinski violations. Default 0.

Details

Uses the generic LogP column with the original threshold of 5.

Value

A data.frame with the rows of data that satisfy all thresholds.

References

Lipinski, C. A., Lombardo, F., Dominy, B. W., & Feeney, P. J. (1997). Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Advanced Drug Delivery Reviews, 23(1-3), 3-25.

Examples

d <- data.frame(MW = 300, LogP = 2, "#H-bond acceptors" = 4,
  "#H-bond donors" = 2, "Lipinski #violations" = 0, check.names = FALSE)
lipinskiFilter(d)

d <- data.frame(MW = 300, LogP = 2, "#H-bond acceptors" = 4,
  "#H-bond donors" = 2, "Lipinski #violations" = 0, check.names = FALSE)
lipinskiFilter(d)

Map user columns to standard ADMET schema

Description

Takes a raw data.frame and a user-specified column mapping, renames columns to the application's standard schema, converts types, optionally calculates missing descriptors with CDK, and computes violation columns and ADMET properties.

Usage

mapADMETColumns(data, mapping, calculate_cdk = TRUE)
mapADMETColumns(data, mapping, calculate_cdk = TRUE)

Arguments

data

A data.frame as uploaded by the user (CSV or Excel).

mapping

A named character vector where names are user column names and values are standard field names. Use "None" for columns to skip. Example: c("mol_weight" = "MW", "logp" = "LogP", "smiles" = "SMILES")

calculate_cdk

Logical. If TRUE and SMILES column is available, calculate missing descriptors (MW, LogP, TPSA, HBD, HBA, RB, MR, HeavyAtoms, AromAtoms) using CDK. Default TRUE.

Details

The mapping argument is a named character vector where each name is the name of a column in data and each value is one of the standard short field codes used by the ADMET Master Manager: "None", "SMILES", "Name", "MW", "LogP", "WLOGP", "TPSA", "HBD", "HBA", "Rotatable Bonds", "Molar Refractivity", "Heavy Atoms", "Aromatic Heavy Atoms", "GI Absorption", "GI Absorption_num", "BBB Permeant", "BBB Permeant_num", "Pgp Substrate", "Pgp Substrate_num", "LogS", "LogD".

Value

A data.frame with standardized column names, violation columns, and ADMET properties.

Map CDK descriptors to the application's standard schema

Description

Translates the raw CDK descriptor names into the application's canonical column names and computes the drug-likeness violation columns and the BOILED-Egg ADMET properties (GI absorption, BBB permeability, P-gp substrate).

Usage

mapCDKDescriptors(cdk_df)
mapCDKDescriptors(cdk_df)

Arguments

cdk_df

A data.frame as returned by calcCDKDescriptors.

Details

The CDK ALogP descriptor is mapped to the generic LogP column. No artificial MLOGP/WLOGP/XLOGP3 columns are created; the application uses a single LogP column for all drug-likeness filters, with thresholds from the original publications.

Value

A data.frame with renamed columns, violation columns and ADMET properties.

Examples


smiles <- c("CCO", "CC(=O)Oc1ccccc1C(=O)O")
desc <- calcCDKDescriptors(smiles)
mapped <- mapCDKDescriptors(desc)

smiles <- c("CCO", "CC(=O)Oc1ccccc1C(=O)O")
desc <- calcCDKDescriptors(smiles)
mapped <- mapCDKDescriptors(desc)

Muegge drug-likeness filter

Description

Filters compounds according to the Muegge pharmacophore-point filter (Muegge et al., 2001): MW, LogP, HBA, HBD, TPSA and rotatable bonds.

Usage

mueggeFilter(
  data,
  m_mw_min = 200,
  m_mw_max = 600,
  m_logp_min = -2,
  m_logp_max = 5,
  m_hba = 10,
  m_hbd = 5,
  m_rb = 15,
  m_tpsa = 150,
  violations = 0
)
mueggeFilter(
  data,
  m_mw_min = 200,
  m_mw_max = 600,
  m_logp_min = -2,
  m_logp_max = 5,
  m_hba = 10,
  m_hbd = 5,
  m_rb = 15,
  m_tpsa = 150,
  violations = 0
)

Arguments

data

A data.frame with the standard column schema.

m_mw_min

Numeric. Minimum molecular weight. Default 200.

m_mw_max

Numeric. Maximum molecular weight. Default 600.

m_logp_min

Numeric. Minimum LogP. Default -2.

m_logp_max

Numeric. Maximum LogP. Default 5.

m_hba

Numeric. Maximum H-bond acceptors. Default 10.

m_hbd

Numeric. Maximum H-bond donors. Default 5.

m_rb

Numeric. Maximum rotatable bonds. Default 15.

m_tpsa

Numeric. Maximum TPSA. Default 150.

violations

Integer. Maximum tolerated Muegge violations. Default 0.

Value

A data.frame with the rows of data that satisfy all thresholds.

References

Muegge, I., Heald, S. L., & Brittelli, D. (2001). Simple selection criteria for drug-like chemical matter. Journal of Medicinal Chemistry, 44(12), 1841-1846.

Examples

d <- data.frame(MW = 300, LogP = 2, TPSA = 80,
  "#H-bond acceptors" = 4, "#H-bond donors" = 2,
  "#Rotatable bonds" = 3, "Muegge #violations" = 0, check.names = FALSE)
mueggeFilter(d)
d <- data.frame(MW = 300, LogP = 2, TPSA = 80,
  "#H-bond acceptors" = 4, "#H-bond donors" = 2,
  "#Rotatable bonds" = 3, "Muegge #violations" = 0, check.names = FALSE)
mueggeFilter(d)

Colour palette selector UI element

Description

A Shiny selectInput that lets the user choose a colour palette for the plots produced by the application.

Usage

palette_selector_ui(id)
palette_selector_ui(id)

Arguments

id

Character. Input id to use for the select input.

Value

A shiny.tag (select input).

Examples

library(shiny)
palette_selector_ui("my_palette")
library(shiny)
palette_selector_ui("my_palette")

BOILED-Egg plot

Description

Produces the BOILED-Egg model plot (Daina & Zoete, 2016) using the official polygon coordinates from the supplementary data (Data S3). Points are coloured by P-gp substrate status when a "Pgp substrate" column is available.

Usage

plotBoiledEgg(data, logp_source = NULL)
plotBoiledEgg(data, logp_source = NULL)

Arguments

data

A data.frame that must contain the columns LogP and TPSA. A "Pgp substrate" column (with values "Yes"/"No") is optional and used for point colouring.

logp_source

Character. Which LogP variant is being used for the BOILED-Egg. If "WLOGP", use the official WLOGP polygons from Daina & Zoete (2016). If any other value (or NULL), use the ALogP-trained polygons. Defaults to auto-detection: if the data has a WLOGP column, WLOGP polygons are used; otherwise ALogP.

Details

The BOILED-Egg model was originally calibrated with WLOGP; the application uses the generic LogP column as an approximation. The axes are TPSA on the x-axis and LogP on the y-axis.

Value

A ggplot2 object.

References

Daina, A., & Zoete, V. (2016). A boiled egg to predict gastrointestinal absorption and brain penetration of small molecules. ChemMedChem, 11(11), 1117-1121.

Examples

d <- data.frame(LogP = c(2, 5), TPSA = c(50, 150))
plotBoiledEgg(d)
d <- data.frame(LogP = c(2, 5), TPSA = c(50, 150))
plotBoiledEgg(d)

Cluster heatmap with dendrogram

Description

Produces a heatmap of physicochemical properties with a dendrogram coupled to the left side (compounds) showing hierarchical clustering. The properties (columns) are also clustered and the column dendrogram appears on top. Values are z-score standardized per property.

Usage

plotClusterHeatmap(
  data,
  variables = NULL,
  id_col = NULL,
  method = "ward.D2",
  scale_data = TRUE,
  palette = "default"
)
plotClusterHeatmap(
  data,
  variables = NULL,
  id_col = NULL,
  method = "ward.D2",
  scale_data = TRUE,
  palette = "default"
)

Arguments

data

A data.frame with physicochemical property columns.

variables

Character vector of numeric column names to use. If NULL, a default set of common descriptors is used.

id_col

Character. Name of the column to use as row labels. If NULL or "None", row numbers are used.

method

Character. Agglomeration method for hierarchical clustering: one of "ward.D", "ward.D2", "single", "complete", "average" (= UPGMA), "mcquitty", "median", "centroid". Default "ward.D2".

scale_data

Logical. Whether to z-score standardize each property before clustering. Default TRUE.

palette

Character. Colour palette name: one of "default", "viridis", "magma", "inferno", "RdBu", "Set1", "Set2". Default "default" (red-blue).

Details

The plot is rendered using base R graphics so it works without additional dependencies beyond what the package already requires.

Value

Invisible NULL; called for the side-effect of drawing the plot on the current graphics device.

References

Murtagh, F., & Contreras, P. (2012). Algorithms for hierarchical clustering: an overview. Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery, 2(1), 86-97.

Examples

d <- data.frame(
  Name = c("A", "B", "C", "D"),
  MW = c(300, 350, 600, 320),
  LogP = c(2, 3, 5, 1),
  TPSA = c(50, 60, 120, 40),
  check.names = FALSE
)
plotClusterHeatmap(d, id_col = "Name")
d <- data.frame(
  Name = c("A", "B", "C", "D"),
  MW = c(300, 350, 600, 320),
  LogP = c(2, 3, 5, 1),
  TPSA = c(50, 60, 120, 40),
  check.names = FALSE
)
plotClusterHeatmap(d, id_col = "Name")

Correlation heatmap of physicochemical properties

Description

Produces a heatmap of the Pearson correlation between key physicochemical properties.

Usage

plotCorrHeatmap(
  data,
  props = c("MW", "LogP", "TPSA", "MR", "#H-bond acceptors", "#H-bond donors",
    "#Rotatable bonds", "#Heavy atoms")
)
plotCorrHeatmap(
  data,
  props = c("MW", "LogP", "TPSA", "MR", "#H-bond acceptors", "#H-bond donors",
    "#Rotatable bonds", "#Heavy atoms")
)

Arguments

data

A data.frame with numeric property columns.

props

Character vector of property column names. Defaults to a set of common descriptors.

Value

A ggplot2 object.

Examples

d <- data.frame(MW = c(300, 400, 500), LogP = c(2, 3, 4),
  TPSA = c(60, 80, 100))
plotCorrHeatmap(d)
d <- data.frame(MW = c(300, 400, 500), LogP = c(2, 3, 4),
  TPSA = c(60, 80, 100))
plotCorrHeatmap(d)

Custom histogram

Description

Produces a highly customizable histogram of any numeric column in the dataset, with optional grouping (overlaid), kernel density overlay, rug plot, and configurable number of bins. This complements the pre-defined plotMW, plotTPSA and plotLogP histograms by allowing the user to pick any numeric variable (including user-supplied columns from CSV uploads, ADMET probabilities, etc.).

Usage

plotHistogramCustom(
  data,
  variable,
  bins = 30,
  group_by = "None",
  show_density = TRUE,
  show_rug = FALSE
)
plotHistogramCustom(
  data,
  variable,
  bins = 30,
  group_by = "None",
  show_density = TRUE,
  show_rug = FALSE
)

Arguments

data

A data.frame.

variable

Character. Name of the numeric variable to histogram.

bins

Integer. Number of bins. Default 30.

group_by

Character. Optional categorical column for overlaid groups, or "None".

show_density

Logical. Overlay a kernel density estimate. Default TRUE.

show_rug

Logical. Show a rug plot at the bottom. Default FALSE.

Value

A ggplot2 object.

Examples

d <- data.frame(MW = c(300, 400, 500, 350, 450, 380),
  group = c("A", "A", "B", "B", "C", "C"))
plotHistogramCustom(d, variable = "MW", group_by = "group")
d <- data.frame(MW = c(300, 400, 500, 350, 450, 380),
  group = c("A", "A", "B", "B", "C", "C"))
plotHistogramCustom(d, variable = "MW", group_by = "group")

LogP distribution histogram

Description

LogP distribution histogram

Usage

plotLogP(data)
plotLogP(data)

Arguments

data

A data.frame containing a numeric LogP column.

Value

A ggplot2 object.

Examples

d <- data.frame(LogP = c(1, 2, 3))
plotLogP(d)
d <- data.frame(LogP = c(1, 2, 3))
plotLogP(d)

Molecular weight distribution histogram

Description

Molecular weight distribution histogram

Usage

plotMW(data)
plotMW(data)

Arguments

data

A data.frame containing a numeric MW column.

Value

A ggplot2 object.

Examples

d <- data.frame(MW = c(300, 400, 500))
plotMW(d)
d <- data.frame(MW = c(300, 400, 500))
plotMW(d)

Parallel coordinates plot

Description

Produces a parallel coordinates plot of the selected numeric variables, optionally coloured by a grouping variable. Requires the suggested package GGally.

Usage

plotParallel(data, variables, color_by = NULL, scale_data = TRUE)
plotParallel(data, variables, color_by = NULL, scale_data = TRUE)

Arguments

data

A data.frame with numeric property columns.

variables

Character vector of numeric column names to display.

color_by

Character. Column to colour lines by, or NULL/ "None" for a single colour.

scale_data

Logical. Whether to standardize variables. Default TRUE.

Value

A ggplot2 object.

Examples


d <- data.frame(MW = c(300, 400, 500), LogP = c(2, 3, 4),
  TPSA = c(60, 80, 100), group = c("A", "A", "B"))
plotParallel(d, variables = c("MW", "LogP", "TPSA"), color_by = "group")

d <- data.frame(MW = c(300, 400, 500), LogP = c(2, 3, 4),
  TPSA = c(60, 80, 100), group = c("A", "A", "B"))
plotParallel(d, variables = c("MW", "LogP", "TPSA"), color_by = "group")

Principal Component Analysis (PCA) chemical space plot

Description

Computes a PCA on the selected numeric physicochemical variables and produces a scatterplot of the first two principal components, optionally coloured by a grouping variable, with 95% confidence ellipses, observation labels and variable loading arrows. Labels require the suggested package ggrepel.

Usage

plotPCA(
  data,
  variables = NULL,
  color_by = "None",
  label_by = "None",
  scale_data = TRUE,
  ellipse = TRUE
)
plotPCA(
  data,
  variables = NULL,
  color_by = "None",
  label_by = "None",
  scale_data = TRUE,
  ellipse = TRUE
)

Arguments

data

A data.frame with numeric property columns.

variables

Character vector of numeric column names to use in the PCA. If NULL, a default set of common descriptors is used.

color_by

Character. Name of the column to colour points by, or "None" for a single colour.

label_by

Character. Name of the column to use as point labels, or "None" for no labels.

scale_data

Logical. Whether to scale variables to unit variance before the PCA. Default TRUE.

ellipse

Logical. Whether to draw 95% confidence ellipses per group. Default TRUE.

Value

A ggplot2 object.

Examples

d <- data.frame(MW = c(300, 400, 500), LogP = c(2, 3, 4),
  TPSA = c(60, 80, 100), group = c("A", "A", "B"))
plotPCA(d, variables = c("MW", "LogP", "TPSA"), color_by = "group")
d <- data.frame(MW = c(300, 400, 500), LogP = c(2, 3, 4),
  TPSA = c(60, 80, 100), group = c("A", "A", "B"))
plotPCA(d, variables = c("MW", "LogP", "TPSA"), color_by = "group")

Radar plot of physicochemical profile

Description

Compares up to 5 molecules simultaneously on a radar chart using 6 key physicochemical properties, normalized 0-1 against the range observed in the full dataset. Requires the suggested package fmsb.

Usage

plotRadar(
  data,
  id_col,
  ids,
  props = c("MW", "LogP", "TPSA", "#H-bond acceptors", "#H-bond donors",
    "#Rotatable bonds")
)
plotRadar(
  data,
  id_col,
  ids,
  props = c("MW", "LogP", "TPSA", "#H-bond acceptors", "#H-bond donors",
    "#Rotatable bonds")
)

Arguments

data

A data.frame with the physicochemical properties.

id_col

Character. Name of the identifier column.

ids

Character vector of molecule identifiers to compare (max 5 recommended).

props

Character vector of property column names to display. Defaults to MW, LogP, TPSA, #H-bond acceptors, #H-bond donors, #Rotatable bonds.

Value

Invisibly NULL; called for the side-effect of drawing the radar chart on the current graphics device.

References

Nakazawa, M. (2019). fmsb: Functions for Medical Statistics Book with some Demographic Data. R package.

Examples


d <- data.frame(
  Name = c("mol1", "mol2"),
  MW = c(300, 400), LogP = c(2, 3), TPSA = c(60, 90),
  "#H-bond acceptors" = c(4, 5), "#H-bond donors" = c(2, 1),
  "#Rotatable bonds" = c(3, 5), check.names = FALSE)
plotRadar(d, id_col = "Name", ids = c("mol1", "mol2"))

d <- data.frame(
  Name = c("mol1", "mol2"),
  MW = c(300, 400), LogP = c(2, 3), TPSA = c(60, 90),
  "#H-bond acceptors" = c(4, 5), "#H-bond donors" = c(2, 1),
  "#Rotatable bonds" = c(3, 5), check.names = FALSE)
plotRadar(d, id_col = "Name", ids = c("mol1", "mol2"))

Tanimoto / AGNES structural similarity dendrogram

Description

Parses SMILES strings, computes extended fingerprints, builds a Tanimoto similarity matrix and clusters the molecules with AGNES. Requires the suggested packages rcdk, fingerprint and cluster, plus a working Java JDK (for rcdk).

Usage

plotTanimoto(
  data,
  smiles_col,
  label_col = NULL,
  max_n = 40,
  method = "average"
)
plotTanimoto(
  data,
  smiles_col,
  label_col = NULL,
  max_n = 40,
  method = "average"
)

Arguments

data

A data.frame containing a SMILES column.

smiles_col

Character. Name of the SMILES column.

label_col

Character. Name of the column to use as dendrogram labels, or NULL to use the SMILES themselves.

max_n

Integer. Maximum number of molecules to include. Default 40.

method

Character. AGNES linking method: one of "average", "complete", "single", "ward". Default "average".

Value

Invisibly the Tanimoto similarity matrix.

References

Willett, P., Barnard, J. M., & Downs, G. M. (1998). Chemical similarity searching. Journal of Chemical Information and Computer Sciences, 38(6), 983-996.

Examples


d <- data.frame(
  Name = c("ethanol", "aspirin", "benzene", "toluene", "phenol"),
  SMILES = c("CCO", "CC(=O)Oc1ccccc1C(=O)O", "c1ccccc1",
            "Cc1ccccc1", "Oc1ccccc1"),
  stringsAsFactors = FALSE)
plotTanimoto(d, smiles_col = "SMILES", label_col = "Name")

d <- data.frame(
  Name = c("ethanol", "aspirin", "benzene", "toluene", "phenol"),
  SMILES = c("CCO", "CC(=O)Oc1ccccc1C(=O)O", "c1ccccc1",
            "Cc1ccccc1", "Oc1ccccc1"),
  stringsAsFactors = FALSE)
plotTanimoto(d, smiles_col = "SMILES", label_col = "Name")

TPSA distribution histogram

Description

TPSA distribution histogram

Usage

plotTPSA(data)
plotTPSA(data)

Arguments

data

A data.frame containing a numeric TPSA column.

Value

A ggplot2 object.

Examples

d <- data.frame(TPSA = c(40, 80, 120))
plotTPSA(d)
d <- data.frame(TPSA = c(40, 80, 120))
plotTPSA(d)

t-SNE chemical space plot

Description

Computes a 2-dimensional t-SNE embedding of the selected numeric physicochemical variables and produces a scatterplot, optionally coloured and labelled. Requires the suggested packages Rtsne and ggrepel.

Usage

plotTSNE(
  data,
  variables,
  color_by = "None",
  label_by = "None",
  perplexity = 30,
  max_iter = 1000,
  scale_data = TRUE
)
plotTSNE(
  data,
  variables,
  color_by = "None",
  label_by = "None",
  perplexity = 30,
  max_iter = 1000,
  scale_data = TRUE
)

Arguments

data

A data.frame with numeric property columns.

variables

Character vector of numeric column names to use.

color_by

Character. Column to colour points by, or "None".

label_by

Character. Column to use as labels, or "None".

perplexity

Numeric. t-SNE perplexity. Default 30.

max_iter

Integer. Number of iterations. Default 1000.

scale_data

Logical. Whether to scale variables. Default TRUE.

Value

A ggplot2 object.

Examples


d <- data.frame(
  MW = rnorm(50, 400, 100),
  LogP = rnorm(50, 3, 1.5),
  TPSA = rnorm(50, 80, 40),
  group = rep(c("A", "B"), each = 25)
)
plotTSNE(d, variables = c("MW", "LogP", "TPSA"),
         color_by = "group", perplexity = 5)

d <- data.frame(
  MW = rnorm(50, 400, 100),
  LogP = rnorm(50, 3, 1.5),
  TPSA = rnorm(50, 80, 40),
  group = rep(c("A", "B"), each = 25)
)
plotTSNE(d, variables = c("MW", "LogP", "TPSA"),
         color_by = "group", perplexity = 5)

UMAP chemical space plot

Description

Produces a 2D UMAP projection of the selected numeric variables for exploratory analysis of the chemical space. Points can be coloured and labelled. Requires the suggested packages uwot and ggrepel.

Usage

plotUMAP(
  data,
  variables,
  color_by = "None",
  label_by = "None",
  n_neighbors = 15,
  min_dist = 0.1,
  scale_data = TRUE
)
plotUMAP(
  data,
  variables,
  color_by = "None",
  label_by = "None",
  n_neighbors = 15,
  min_dist = 0.1,
  scale_data = TRUE
)

Arguments

data

A data.frame with numeric property columns.

variables

Character vector of numeric column names to use.

color_by

Character. Column to colour points by, or "None".

label_by

Character. Column to use as labels, or "None".

n_neighbors

Integer. Number of nearest neighbours used by UMAP. Default 15.

min_dist

Numeric. Minimum distance between points in the embedding. Default 0.1.

scale_data

Logical. Whether to scale variables. Default TRUE.

Value

A ggplot2 object.

Examples


d <- data.frame(MW = c(300, 400, 500,600), LogP = c(2, 3, 4,1),
  TPSA = c(60, 80, 100,40), group = c("A", "A", "B", "B"))
plotUMAP(d, variables = c("MW", "LogP", "TPSA"), color_by = "group")

d <- data.frame(MW = c(300, 400, 500,600), LogP = c(2, 3, 4,1),
  TPSA = c(60, 80, 100,40), group = c("A", "A", "B", "B"))
plotUMAP(d, variables = c("MW", "LogP", "TPSA"), color_by = "group")

Violin plot

Description

Produces a violin plot of a numeric variable grouped by a categorical variable, with an optional inner boxplot and jittered points.

Usage

plotViolin(data, variable, group_by, show_box = TRUE, show_points = FALSE)
plotViolin(data, variable, group_by, show_box = TRUE, show_points = FALSE)

Arguments

data

A data.frame.

variable

Character. Name of the numeric variable to plot.

group_by

Character. Name of the categorical grouping variable.

show_box

Logical. Whether to overlay a boxplot. Default TRUE.

show_points

Logical. Whether to overlay jittered points. Default FALSE.

Value

A ggplot2 object.

Examples

d <- data.frame(MW = c(300, 400, 500),
  group = c("A", "A", "B"))
plotViolin(d, variable = "MW", group_by = "group")
d <- data.frame(MW = c(300, 400, 500),
  group = c("A", "A", "B"))
plotViolin(d, variable = "MW", group_by = "group")

Launch the ADMETShiny application

Description

Starts the interactive ADMETShiny Shiny application. The function registers the package's static web assets (the hex sticker) and returns a Shiny app object that can be printed or run directly.

Usage

run_app(onStart = NULL, ...)
run_app(onStart = NULL, ...)

Arguments

onStart

An optional function called once when the app starts (forwarded to shiny::shinyApp).

...

Additional arguments forwarded to shiny::shinyApp (e.g. options).

Value

A shiny.appobj object (invisibly); called for the side-effect of launching a Shiny application when printed.

Examples

if (interactive()) {
  admetshiny::run_app()
}

if (interactive()) {
  admetshiny::run_app()
}

Veber drug-likeness filter

Description

Filters compounds according to the Veber oral-bioavailability rule (Veber et al., 2002): rotatable bonds and a polarity condition (TPSA or HBA+HBD).

Usage

veberFilter(data, v_rb = 10, v_tpsa = 140, v_hb_sum = 12, violations = 0)
veberFilter(data, v_rb = 10, v_tpsa = 140, v_hb_sum = 12, violations = 0)

Arguments

data

A data.frame with the standard column schema.

v_rb

Numeric. Maximum rotatable bonds. Default 10.

v_tpsa

Numeric. Maximum TPSA. Default 140.

v_hb_sum

Numeric. Maximum H-bond acceptors + donors. Default 12.

violations

Integer. Maximum tolerated Veber violations. Default 0.

Value

A data.frame with the rows of data that satisfy all thresholds.

References

Veber, D. F., Johnson, S. R., Cheng, H. Y., Smith, B. R., Ward, K. W., & Kopple, K. D. (2002). Molecular properties that influence the oral bioavailability of drug candidates. Journal of Medicinal Chemistry, 45(12), 2615-2623.

Examples

d <- data.frame(TPSA = 80, "#Rotatable bonds" = 3,
  "#H-bond acceptors" = 4, "#H-bond donors" = 2,
  "Veber #violations" = 0, check.names = FALSE)
veberFilter(d)
d <- data.frame(TPSA = 80, "#Rotatable bonds" = 3,
  "#H-bond acceptors" = 4, "#H-bond donors" = 2,
  "Veber #violations" = 0, check.names = FALSE)
veberFilter(d)

Package 'admetshiny'

Help Index

admetshiny: Interactive ADMET and Drug-Likeness Analysis

Description

Details

Drug-likeness filters

Data normalization

CDK & webchem

Visualizations

Application

Author(s)

See Also

Apply a colour palette to a ggplot object

Description

Usage

Arguments

Value

Examples

Apply drug-likeness filters in sequence

Description

Usage

Arguments

Value

See Also

Examples

Calculate molecular descriptors with CDK

Description

Usage

Arguments

Value

See Also

Examples

Compute BOILED-Egg ADMET properties

Description

Usage

Arguments

Details

Value

References

Examples

Compute drug-likeness violation columns

Description

Usage

Arguments

Details

Value

See Also

Examples

Egan drug-likeness filter

Description

Usage

Arguments

Value

References

Examples

Generate an ADMETShiny report from the R console

Description

Usage

Arguments

Value

Examples

Retrieve canonical SMILES from PubChem

Description

Usage

Arguments

Value

See Also

Examples

Ghose drug-likeness filter

Description

Usage

Arguments

Details

Value

References

Examples

Lipinski Rule-of-Five filter

Description

Usage

Arguments