desc2fp: Fingerprints from descriptor vectors

Description Usage Arguments Details Value Author(s) References See Also Examples

Description

Generates fingerprints from descriptor vectors such as atom pairs stored in APset or list containers. The obtained fingerprints can be used for structure similarity comparisons, searching and clustering. Due to their compact size, computations on fingerprints are often more time and memory efficient than on their much more complex atom pair counterparts.

Usage

1
desc2fp(x, descnames=1024, type = "FPset")

Arguments

x

Object of classe APset or list of vectors

descnames

Descriptor set to consider for fingerprint encoding. If a single value from 1-4096 is provided then the function uses the corresponding number of the most frequent atom pairs stored in the apfp data set provided by the package. Alternatively, one can provide here any custom atom pair selection in form of a character vector.

type

return fingerprint set as FPset, matrix or character vector

Details

...

Value

matrix or character vectors

Author(s)

Thomas Girke

References

Chen X and Reynolds CH (2002). "Performance of similarity measures in 2D fragment-based similarity searching: comparison of structural descriptors and similarity coefficients", J Chem Inf Comput Sci.

See Also

Functions: sdf2ap, SDF2apcmp, apset2descdb, cmp.search, cmp.similarity

Related classes: SDF, SDFset, SDFstr, APset.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
## Instance of SDFset class
data(sdfsample)
sdfset <- sdfsample[1:10]

## Compute atom pair library
apset <- sdf2ap(sdfset)

## Compute atom pair fingerprint matrix using internal atom pair
## selection containing 4096 most common atom pairs in DrugBank.
## For details see ?apfp. The following example uses from this 
## set the 1024 most frequent atom pairs: 
fpset <- desc2fp(x=apset, descnames=1024, type="FPset")

## Alternatively, one can provide any custom atom pair selection. Here
## 1024 most common ones in apset object.
fpset1024 <- names(rev(sort(table(unlist(as(apset, "list")))))[1:1024])
fpset2 <- desc2fp(x=apset, descnames=fpset1024, type="FPset")

## A more compact way of storing fingerprints is as character values
fpchar <- desc2fp(x=apset, descnames=1024, type="character")

## Convert character fingerprints back to FPset or matrix
fpset <- as(fpchar, "FPset")
fpma <- as.matrix(fpset)

## Similarity searching returning Tanimoto similarity coefficients
fpSim(x=fpset[1], y=fpset)

## Clustering example
simMAap <- sapply(cid(fpset), function(x) fpSim(x=fpset[x], fpset, sorted=FALSE)) 
hc <- hclust(as.dist(1-simMAap), method="single")
plot(as.dendrogram(hc), edgePar=list(col=4, lwd=2), horiz=TRUE)

girke-lab/ChemmineR-git-svn-bridge documentation built on May 17, 2019, 5:25 a.m.