This pharmaceutical data set was used to develop a model for predicting compounds' permeability. In short, permeability is the measure of a molecule's ability to cross a membrane. The body, for example, has notable membranes between the body and brain, known as the blood-brain barrier, and between the gut and body in the intestines. These membranes help the body guard critical regions from receiving undesirable or detrimental substances. For an orally taken drug to be effective in the brain, it first must pass through the intestinal wall and then must pass through the blood-brain barrier in order to be present for the desired neurological target. Therefore, a compound's ability to permeate relevant biological membranes is critically important to understand early in the drug discovery process. Compounds that appear to be effective for a particular disease in research screening experiments, but appear to be poorly permeable may need to be altered in order improve permeability, and thus the compound's ability to reach the desired target. Identifying permeability problems can help guide chemists towards better molecules.
Permeability assays such as PAMPA and Caco-2 have been developed to help measure compounds' permeability (Kansy et al, 1998). These screens are effective at quantifying a compound's permeability, but the assay is expensive labor intensive. Given a sufficient number of compounds that have been screened, we could develop a predictive model for permeability in an attempt to potentially reduce the need for the assay. In this project there were 165 unique compounds; 1107 molecular fingerprints were determined for each. A molecular fingerprint is a binary sequence of numbers that represents the presence or absence of a specific molecular sub-structure. The response is highly skewed, the predictors are sparse (15.5 percent are present), and many predictors are strongly associated.
Usage
data(permeability)
Value
permeability
permeability values for each compound.
fingerprints
a matrix of binary fingerprint indicator variables.
Source
Kansy, M., Senner, F., and Gubernator, K. (1998). Physicochemical High Throughput Screening: Parallel Artificial Membrane Permeation Assay in the Description of Passive Absorption Processes. J. Med. Chem, 41(7), 1007-1010.
R version 3.3.1 (2016-06-21) -- "Bug in Your Hair"
Copyright (C) 2016 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)
R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.
R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.
Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.
> library(AppliedPredictiveModeling)
> png(filename="/home/ddbj/snapshot/RGM3/R_CC/result/AppliedPredictiveModeling/permeability.Rd_%03d_medium.png", width=480, height=480)
> ### Name: permeability
> ### Title: Permeability Data
> ### Aliases: permeability fingerprints
> ### Keywords: datasets
>
> ### ** Examples
>
> data(permeability)
>
> hist(permeability)
>
> summary(apply(fingerprints, 2, mean))
Min. 1st Qu. Median Mean 3rd Qu. Max.
0.000000 0.006061 0.024240 0.154800 0.181800 1.000000
>
>
>
>
>
> dev.off()
null device
1
>
providing 
.