Package 'fmcsR'

A FMCS solver package.

Description

The package consists of two main functions, fmcs which computes the flexible MCS between two SDF objects. And fmcsBatch runs the FMCS algorithm on a SDFset.

Details

Package:	fmcsR
Type:	Package
Version:	1.0
Date:	2012-02-01

Author(s)

Yan Wang

Maintainer: Yan Wang <[email protected]>

Examples

library(fmcsR)
data(sdfsample)
sdfset <- sdfsample
result1 <- fmcs(sdfset[[1]], sdfset[[2]])
result2 <- fmcs(sdfset[[1]], sdfset[[2]], au=3)
result3 <- fmcs(sdfset[[1]], sdfset[[2]], bu=3)
result4 <- fmcs(sdfset[[1]], sdfset[[2]], au=1, bu=1)
result5 <- fmcs(sdfset[[1]], sdfset[[2]], matching.mode="aromatic")
result6 <- fmcs(sdfset[[1]], sdfset[[2]], au=2, bu=1, matching.mode="aromatic")

fmcsBatch(sdfset[[1]], sdfset[1:3])
fmcsBatch(sdfset[[1]], sdfset[1:3], au=2)
fmcsBatch(sdfset[[1]], sdfset[1:3], bu=1)
fmcsBatch(sdfset[[1]], sdfset[1:3], matching.mode="aromatic", au=1, bu=1)

---

Flexible MCS (FMCS) Finder

Description

R function to call the C++ implementation of the flexible common substructure (FMCS) algorithm. The FMCS algorithm provides an improved maximum common substructure (MCS) search method that allows atom and/or bond mismatches in the substructures shared among two small molecules. The resulting flexible MCSs (FMCSs) are often larger than strict MCSs, resulting in the identification of more common features in their source structures, as well as a higher sensitivity in detecting weak similarities among compounds.

Usage

fmcs(sdf1, sdf2, al = 0, au = 0, bl = 0, bu = 0, matching.mode =
	  "static", fast = FALSE, timeout=60000)

Arguments

sdf1	Input query SDF object or SDFset object with a single molecule.
sdf2	Input target SDF object SDFset object with a single molecul.
al	Lower bound for the number of atom mismatches.
au	Upper bound for the number of atom mismatches.
bl	Lower bound for the number of bond mismatches.
bu	Upper bound for the number of bond mismatches.
matching.mode	Three modes for bond matching are supported: "static", "aromatic", and "ring".
fast	If fast is set to TURE, then the fast computing mode will be turned on. In this case, the algorithm will only return the size information about the source structures and their MCSs, while omitting all structural information.
timeout	The maximum amount of time to spend searching, in milliseconds. A value of 0 indicates no timeout.

Details

...

Value

Returns object of class MCS

Author(s)

Yan Wang, Thomas Girke

References

Publication in preparation.

See Also

plotMCS, fmcsBatch, ?"MCS-class"

Examples

library(fmcsR)
data(sdfsample)
sdfset <- sdfsample
mcs1 <- fmcs(sdfset[[1]], sdfset[[2]])
mcsfast <- fmcs(sdfset[[1]], sdfset[[2]], fast=TRUE)
mcs2 <- fmcs(sdfset[[1]], sdfset[[2]], au=3)
mcs3 <- fmcs(sdfset[[1]], sdfset[[2]], bu=3)
mcs4 <- fmcs(sdfset[[1]], sdfset[[2]], au=1, bu=1)
mcs5 <- fmcs(sdfset[[1]], sdfset[[2]], matching.mode="aromatic")
mcs6 <- fmcs(sdfset[[1]], sdfset[[2]], au=2, bu=1, matching.mode="aromatic")

## Plot MCS objects
plotMCS(mcs6)

## Methods to return components of MCS objects
stats(mcs6)
mcs6[["stats"]]
mcs1(mcs6)
mcs6[["mcs1"]]
mcs2(mcs6)
mcs6[["mcs2"]]

## Constructor method from list
mylist <- list(stats=stats(mcs6), mcs1=mcs1(mcs6), mcs2=mcs2(mcs6))
mymcs <- as(mylist, "MCS")

---

FMCS Search Function

Description

Compound search function that runs the FMCS algorithm for a query compound against a set of molecules stored in an SDFset container.

Usage

fmcsBatch(querySdf, sdfset, al = 0, au = 0, bl = 0, bu = 0,
matching.mode = "static",timeout=60000,numParallel=1)

Arguments

querySdf	Input query SDF object or SDFset object of length one.
sdfset	Input target SDFset object.
al	Lower bound for the number of atom mismatches.
au	Upper bound for the number of atom mismatches.
bl	Lower bound for the number of bond mismatches.
bu	Upper bound for the number of bond mismatches.
matching.mode	Three matching mode are supported, "static", "aromatic", and "ring".
timeout	The maximum amount of time to spend on each pair of comparisons, in milliseconds. A value of 0 indicates no timeout.
numParallel	The number of comparisons to run in parallel, using local cores.

Details

This function runs the FMCS algorithm in fast computing mode. Thus, it will only return the similarity scores and size information about the source structures and their MCSs, while omitting all structural information.

Value

Returns a matrix with compound IDs as row names and the following columns: Query_Size, Target_Size, MCS_Size, Tanimoto_Coefficient and Overlap_Coefficient. For details see vignette of this package.

Author(s)

Yan Wang, Thomas Girke

See Also

plotMCS, fmcs, ?"MCS-class"

Examples

library(fmcsR)
data(sdfsample)
sdfset <- sdfsample
fmcsBatch(sdfset[[1]], sdfset[1:3])
fmcsBatch(sdfset[[1]], sdfset[1:3], au=2)
fmcsBatch(sdfset[[1]], sdfset[1:3], bu=1)
fmcsBatch(sdfset[[1]], sdfset[1:3], matching.mode="aromatic", au=1, bu=1)

---

SD file stored in SDFset object

Description

Sample compound structures stored in SDF format.

Usage

data(fmcstest)

Format

Object of class SDFset

Details

Object stores X molecules from a sample SD file.

Source

ftp://ftp.ncbi.nih.gov/pubchem/

References

SDF format definition: http://www.symyx.com/downloads/public/ctfile/ctfile.jsp

Examples

data(fmcstest)
sdfset <- fmcstest
view(sdfset)

---

Class "MCS"

Description

List-like container for storing results from fmcs function.

Objects from the Class

Objects can be created by calls of the form new("MCS", ...).

Slots

stats:

Object of class "numeric" ~~

mcs1:

Object of class "SDFset" ~~

mcs2:

Object of class "SDFset" ~~

Methods

[[

signature(x = "MCS"): ...

coerce

signature(from = "list", to = "MCS"): ...

mcs1

signature(x = "MCS"): ...

mcs2

signature(x = "MCS"): ...

stats

signature(x = "MCS"): ...

Note

...

Author(s)

Yan Wang

References

...

See Also

Related classes: SDF, SDFstr

Examples

## Create MCS instance
showClass("MCS")
data(sdfsample)
sdfset <- sdfsample
mcs <- fmcs(sdfset[[1]], sdfset[[2]], au=2, bu=2)

## Methods to return components of MCS 
stats(mcs)
mcs[["stats"]]
mcs1(mcs)
mcs[["mcs1"]]
mcs2(mcs)
mcs[["mcs2"]]

## Constructor method from list
mylist <- list(stats=stats(mcs), mcs1=mcs1(mcs), mcs2=mcs2(mcs))
mymcs <- as(mylist, "MCS")

---

Return MCS object as SDFset

Description

Helper function to run atomsubset from ChemmineR library on MCS objects in order to obtain their results in SDFset format.

Usage

mcs2sdfset(x, ...)

Arguments

x	Object of class MCS
...	Arguments to be passed to/from other methods.

Details

Returns MCS data in form of a list containing two SDFset objects, one for the query and one for the target structure.

Value

List with two SDFset objects.

Note

...

Author(s)

Thomas Girke

References

...

See Also

fmcs

Examples

library(fmcsR)
data(sdfsample)
sdfset <- sdfsample
mcs <- fmcs(sdfset[[1]], sdfset[[2]], au=2, bu=1, matching.mode="aromatic")
mcs2sdfset(x=mcs, type="new")
mcs2sdfset(x=mcs, type="old")[[1]][[1]] 
plot(mcs2sdfset(x=mcs, type="new")[[1]][1])

---

Plot MCS

Description

Convenience plotting function to visualize and compare MCSs generated by fmcs function.

Usage

plotMCS(x, mcs = 1, print = FALSE, ...)

Arguments

x	MCS object
mcs	Selection of MCS solution by position number, default is 1.
print	print=FALSE turns of printing behavior of class.
...	Arguments to be passed to/from other methods.

Details

The two structures, target and query, used to generate x with a call to fmcs are plotted next to each other, and the corresponding MCS substructures are highlighted in color.

Value

Prints summary of MCS to screen and plots their structures to graphics device.

Note

...

Author(s)

Yan Wang

References

...

See Also

sdf.visualize

Examples

library(fmcsR)
data(sdfsample)
sdfset <- sdfsample
mcs <- fmcs(sdfset[[1]], sdfset[[2]], au=2, bu=1, matching.mode="aromatic")
plotMCS(mcs, mcs=1)

---