frags: Molecule Fragmentation Methods

Description Usage Arguments Value Author(s) See Also Examples


A variety of methods for fragmenting molecules are available ranging from exhaustive, rings to more specific methods such as Murcko frameworks. Fragmenting a collection of molecules can be a useful for a variety of analyses. In addition fragment based analysis can be a useful and faster alternative to traditional clustering of the whole collection, especially when it is large.

Note that exhaustive fragmentation of large molecules (with many single bonds) can become time consuming.


get.murcko.fragments(mols, min.frag.size = 6, as.smiles = TRUE, single.framework = FALSE)
get.exhaustive.fragments(mols, min.frag.size = 6, as.smiles = TRUE)



A molecule object or list of molecule objects. Each object should have a jclass of IAtomContainer


The size of the smallest fragments to be considered


If TRUE, the fragments are returned as SMILES strings, otherwise as IAtomContainer objects


If TRUE, then a single framework (i.e., the framework consisting of the union of all ring systems and linkers) is returned for each molecule. Otherwise, all combinations of ring systems and linkers are returned


get.murcko.fragments returns a list with each element being a list with two elements: rings and frameworks. Each of these elements is either a character vector of SMILES strings or a list of IAtomContainer objects. get.exhaustive.fragments returns a list of length equal to the number of input molecules. Each element is a character vector of SMILES strings or a list of IAtomContainer objects.


Rajarshi Guha (

See Also

load.molecules, parse.smiles,


mol <- parse.smiles('c1ccc(cc1)CN(c2cc(ccc2[N+](=O)[O-])c3c(nc(nc3CC)N)N)C')[[1]]
mf1 <- get.murcko.fragments(mol, as.smiles=TRUE, single.framework=TRUE)
mf1 <- get.murcko.fragments(mol, as.smiles=TRUE, single.framework=FALSE)

Search within the rcdk package
Search all R packages, documentation and source code

Questions? Problems? Suggestions? or email at

Please suggest features or report bugs with the GitHub issue tracker.

All documentation is copyright its authors; we didn't write any of that.