/rcdk/man/getfp.Rd
Unknown | 87 lines | 79 code | 8 blank | 0 comment | 0 complexity | 413d6e3f5378e2ff04e77c0b26f4899e MD5 | raw file
1\name{get.fingerprint} 2\alias{get.fingerprint} 3\title{ 4 Evaluate Fingerprints 5} 6\description{ 7 This function evaluates fingerprints of a specified type for a set of molecules or 8 a single molecule. Depending on the nature of the fingerprint, 9 parameters can be specified. Currently five different fingerprints 10 can be specified: 11 \itemize{ 12 \item standard - Considers paths of a given length. The default is 13 but can be changed. These are hashed fingerprints, with a 14 default length of 1024 15 \item extended - Similar to the standard type, but takes rings and 16 atomic properties into account into account 17 \item graph - Similar to the standard type by simply considers connectivity 18 \item hybridization - Similar to the standard type, but only consider hybridization state 19 \item maccs - The popular 166 bit MACCS keys described by MDL 20 \item estate - 79 bit fingerprints corresponding to the E-State atom types described by Hall and Kier 21 \item pubchem - 881 bit fingerprints defined by PubChem 22 \item kr - 4860 bit fingerprint defined by Klekota and Roth 23 \item shortestpath - A fingerprint based on the shortest paths between pairs of atoms and takes into account ring systems, charges etc. 24 \item signature - A feature,count type of fingerprint, similar in nature to circular fingerprints, but based on the signature 25 descriptor 26 \item circular - An implementation of the ECFP6 fingerprint 27 } 28 29 Depending on whether the input is a single \code{IAtomContainer} 30 object, a list or single vector is returned. Each element of the list 31 is an S4 object of class \code{\link{fingerprint-class}} or \code{\link{featvec-class}}, 32 which can be manipulated with the fingerprint package. 33} 34\usage{ 35 get.fingerprint(molecule, type = 'standard', 36 fp.mode = 'bit', depth=6, size=1024, verbose=FALSE) 37} 38\arguments{ 39 \item{molecule}{An \code{IAtomContainer} object that can be obtained 40 by loading them from disk or drawing them in the editor. 41 } 42 \item{type}{The type of fingerprint. See description for possible values. The default is the standard 43 binary fingerprint.} 44 \item{fp.mode}{The type of fingerprint to return. Possible values are 'bit', 'raw', and 'count'. The 'raw' mode will 45 return a \code{featvec-class} type of fingerprint, representing fragments and their count of occurence in the molecule. 46 The 'count' mode is similar, except that it returns hash values of fragments and their count of occurence. 47 While any of these values can be specified, a given fingerprint implementation may not implement all of them, and in 48 those cases the return value is NULL.} 49 \item{depth}{The search depth. This argument is ignored for the 50 'pubchem', 'maccs', 'kr' and 'estate' fingerprints} 51 \item{size}{The length of the fingerprint bit string. This argument is ignored for the 52 'pubchem', 'maccs', 'kr', 'signature', 'circular' and 'estate' fingerprints} 53 \item{verbose}{If \code{TRUE}, exceptions, if they occur, will be printed} 54} 55\value{ 56 Objects of class \code{\link{fingerprint-class}} or \code{\link{featvec-class}}, from the \code{fingerprint} package. If there is a problem 57 during fingerprint calculation, \code{NULL} is returned. 58} 59\examples{ 60## get some molecules 61sp <- get.smiles.parser() 62smiles <- c('CCC', 'CCN', 'CCN(C)(C)', 'c1ccccc1Cc1ccccc1','C1CCC1CC(CN(C)(C))CC(=O)CC') 63mols <- parse.smiles(smiles) 64 65## get a single fingerprint using the standard 66## (hashed, path based) fingerprinter 67fp <- get.fingerprint(mols[[1]]) 68 69## get MACCS keys for all the molecules 70fps <- lapply(mols, get.fingerprint, type='maccs') 71 72## get Signature fingerprint 73## feature, count fingerprinter 74fps <- lapply(mols, get.fingerprint, type='signature', fp.mode='raw') 75} 76 77\seealso{ 78 \code{\link{load.molecules}} 79} 80 81\references{ 82Faulon et al, \href{http://dx.doi.org/10.1021/ci020345w}{The Signature Molecular Descriptor. 1. Using Extended Valence Sequences in QSAR and QSPR studies}, \emph{J. Chem. Inf. Comput. Sci.}, 2003, \emph{43}, 707-720. 83} 84 85\keyword{programming} 86 87\author{Rajarshi Guha (\email{rajarshi.guha@gmail.com})}