RMol: a toolset for transforming SD/Molfile structure information into R objects
 Martin Grabner^{1},
 Kurt Varmuza^{2} and
 Matthias Dehmer^{1}Email author
https://doi.org/10.1186/17510473712
© Grabner et al; licensee BioMed Central Ltd. 2012
Received: 27 September 2012
Accepted: 30 October 2012
Published: 14 November 2012
Abstract
Background
The graphtheoretical analysis of molecular networks has a long tradition in chemoinformatics. As demonstrated frequently, a well designed format to encode chemical structures and structurerelated information of organic compounds is the Molfile format. But when it comes to use modern programming languages for statistical data analysis in Bio and Chemoinformatics, R as one of the most powerful free languages lacks tools to process Molfile data collections and import molecular network data into R.
Results
We design an R object which allows a lossless information mapping of structural information from Molfiles into R objects. This provides the basis to use the RMol object as an anchor for connecting Molfile data collections with R libraries for analyzing graphs. Associated with the RMol objects, a set of R functions completes the toolset to organize, describe and manipulate the converted data sets. Further, we bypass Rtypical limits for manipulating large data sets by storing R objects in bzcompressed serialized files instead of employing RData files.
Conclusions
By design, RMol is a R toolset without dependencies to other libraries or programming languages. It is useful to integrate into pipelines for serialized batch analysis by using network data and, therefore, helps to process sdfdata sets in R efficiently. It is freely available under the BSD licence. The script source can be downloaded fromhttp://sourceforge.net/p/rmoltoolset
Background
To solve many tasks in Bio and Chemoinformatics, the analysis of chemical and biological structures represented by networks has been proven powerful[1, 2]. A typical problem in this area is to characterize the structure of molecular networks quantitatively by using graph measures[3–7] or to predict physicochemical properties of the molecules by taking structural features into account[8].
For quantifying structural information of molecular networks, one often needs quantitative or comparative network measures to analyze the structure of the underlying networks[1, 2]. For instance, Dragon[9] is a commercial and wellknown software to calculate socalled molecular descriptors from SD/Molfile data[10] and other data formats specializing in chemical structures. But when using the programming environment R[11], there is yet no interface to employ structural information of molecular networks encoded by SD/Molfile data.
To tackle this problem and, hence, to spread out the usage of R in chemically and biologicallydriven disciplines, we develop an R toolset for transforming SD/Molfile structure information into R objects. As structural information of the networks is now available in R, we hope that our tool may stimulate the Bioinformatics community to explore problems centered around chemical and molecular networks by using existing R packages.
Tools for graph analysis
In this section, we briefly sketch some tools for analyzing graphs by using R. An extensive review of such tools can be found in[12, 13].
Among other environments suitable for graph analysis[14], the script language R has gained much importance. Not only because basic functions are allocated by R packages such as the packages graph[15] from Bioconductor[16] and igraph[17] but also because of packages such as QuACN[18]. The latter contributed extensively to analyze networks quantitatively[18] with R. Note that QuACN is an Rtool for calculating ca. 150 quantitative network measures which can be mostly interpreted as complexity indices[19].
Also, Guha[20] developed a set of wrapper functions providing R user access to the functions and objects of the CDK[21] representing a Java framework for cheminformatics. The chemoinformatics package ChemMineR[22] written in R includes in updated versions functions which are capable of reading and extracting structural information from different data formats including SDfiles. But we emphasize that both packages are conceptionally focused on the inspection of single networks or comparison of limited data sets, in contrast to the RMol script collection which fits well into workflows with serialized pipelines. To reduce dependencies, we extract SDfile information with our own R parser. Also, to avoid repetitive transformation processes we store the chemical and molecular network information in R objects.
Results and discussion
Besides the definition of the novel RMol object for encoding chemical structure information by using R, we develop a set of functions for the programming environment R to accomplish and facilitate efficient graphbased analysis of the underlying molecular networks, e.g., chemical structures. More precisely, this tool covers the following functionalities:

Importing chemical structure data from an SDfile (Molfile format) into R.

Handling of RMol data sets as serialized bzcompressed files (to bypass memory limits).

Providing simple statistics of chemical structures or structure data sets in RMol format.

A filter for selecting chemical structures and reorganizing data collections in RMol format.

Generating adjacency matrices or connection tables from chemical structures in RMol format.

Converting RMol objects into attributeextended graphNEL objects. By doing so, this links directly to R packages for graph analysis (e.g., see graph, igraph, QuACN).
In the following, we explain some items of RMol in more detail. The function Sdf2RMol has been developed to process SDfiles and convert chemical structure information from the Molfile portions into RMol objects.
Concretely Sdf2RMol represents a working script, which combines an entry picking routine (pickSdfEntry) with an RMol specific parser (parseSdfEntry) using regular expressions to scan the Molfile sections of SDfiles according to the CTfile format specifications[23]. Moreover Sdf2RMol completes the conversion pipeline with error logging and internal routines for checking feeded entries for consistency and plausibility. Finally the resulting R objects are streamed as data sets into serialized bzcompressed files.
These files are denoted with the file ending .Rbz and referred to as ”Rbzfiles”. Rbzfiles help to bypass Rtypical memory limits for huge data collections and are useful storage containers for any R object. By design, Rbzfiles contain the R objects as serialized list elements S[i], where S[i]=list(objectname[i],objectcontent[i]). They are a useful data source for any R driven process pipeline.
The functions RData2Rbz and Rbz2RData allow the transformation of the serial Rbzformat to standard RDataformat and vice versa. For users who are not familiar with connection manipulation in R, we also include the functions RbzOpen, NextRbzObject, RbzClose to alleviate the handling of Rbzfiles.
To extract and summarize properties of Rbzpacked RMol data sets, the functions RbzSummary and RbzSummaryReport are useful. To manipulate and split these data sets RbzFilter is available.
The raw graphNEL class, as defined in the R package graph is sufficient to build representations for graphs without vertex and edge labels. However, to perform the analysis of labeled graphs (e.g., graphs representing chemical structures with hetero atoms and bond types) by using the QuACN package, graphNEL needs to be extended with bond and atom attributes. RMol contains the function RMol2QuACNgN to pack the relevant information into these attributeextended graphNEL objects. All RMol functions are put together in one R script. After sourcing this script all functions will be available to support in preparing chemical structure data for analyzing molecular networks.
Conclusions
In this report, we presented an R toolset to convert the structure information of molecular graphs encoded by SD/Molfiles into R objects. It complements existing packages capable of reading SDfile information by easying batch processing and using pure R scripts without dependencies. In combination with R packages designed for analysing graph and network properties it represents a connector module for R workflows which process structure information from SDfiles. This toolset can also support other R packages for analyzing networks structurally and, thus, makes a further contribution towards demonstrating the power of R for network analysis in Chemo and Bioinformatics.
So far, it was not common to investigate SDfile data collections by using R and packages thereof. The new toolset RMol may encourage the community to spread out the usage of R for chemically and biologically driven areas.
Declarations
Acknowledgements
Funding
Martin Grabner, Kurt Varmuza and Matthias Dehmer thank the Austrian Science Funds for supporting this work (project P22029N13).
Authors’ Affiliations
References
 EmmertStreib F, Dehmer M: Networks for systems biology: conceptual connection of data and function. IET Syst Biol. 2011, 5: 185207. 10.1049/ietsyb.2010.0025.View ArticlePubMedGoogle Scholar
 Gasteiger J, Engel T: Chemoinformatics  A Textbook. 2003, Weinheim, Germany: Wiley VCHView ArticleGoogle Scholar
 Bonchev D: Information Theoretic Indices for Characterization of Chemical Structures. 1983, Chichester: Research Studies PressGoogle Scholar
 Dehmer M, Varmuza K, Borgert S, EmmertStreib F: On entropybased molecular descriptors: statistical analysis of real and synthetic chemical structures. J Chem Inf Model. 2009, 49: 16551663. 10.1021/ci900060x.View ArticlePubMedGoogle Scholar
 Dehmer M, Barbarini N, Varmuza K, Graber A: A large scale analysis of informationtheoretic network complexity measures using chemical structures. PLoS ONE. 2009, 4 (12): e805710.1371/journal.pone.0008057.PubMed CentralView ArticlePubMedGoogle Scholar
 Dehmer M, Barbarini N, Varmuza K, Graber A: Novel topological descriptors for analyzing biological networks. BMC Struct Biol. 2010, 10: 1810.1186/147268071018.PubMed CentralView ArticlePubMedGoogle Scholar
 Todeschini R, Consonni V, Mannhold R: Handbook of Molecular Descriptors. 2002, Weinheim, Germany: WileyVCHGoogle Scholar
 Devillers J, Balaban AT: Topological Indices and Related Descriptors in QSAR and QSPR. 1999, Amsterdam, The Netherlands: Gordon and Breach Science PublishersGoogle Scholar
 Todeschini R, Consonni V, Mauri A, Pavan M: Dragon, software for calculation of molecular descriptors. 2004, Milano, Italy: Talete srlGoogle Scholar
 Dalby A, Nourse JG, Hounshell WD, Gushurst AKI, Grier DL, Leland BA, Laufer J: Description of several chemical structure file formats used by computer programs developed at Molecular Design Limited. J Chem Inf Comput Sci. 1992, 32 (3): 244255. 10.1021/ci00007a012.View ArticleGoogle Scholar
 R Development Core Team: R: A Language and Environment for Statistical Computing. 2008, Vienna, Austria: R Foundation for Statistical Computing, [http://www.Rproject.org] [ISBN 3900051070]..Google Scholar
 Meur NL, Gentleman R: Analyzing biological data using R: methods for graphs and networks. Methods Mol Biol. 2012, 804: 343373. 10.1007/9781617793615_19.View ArticlePubMedGoogle Scholar
 Müller L, Dehmer M, EmmertStreib F: Networkbased Methods for Computational Diagnostics by Means of R. Computational Medicine. Edited by: Trajanoski Z. 2012, Vienna, Austria: Springer, 185197.View ArticleGoogle Scholar
 Hagberg AA, Schult DA, Swart PJ: Exploring network structure, dynamics, and function using NetworkX. Proceedings of the 7th Python in Science Conference, SciPy2008. 2008, 1115.Google Scholar
 Gentleman R, Whalen E, Huber W, Falcon S: graph: A package to handle graph data structures. 2010, [http://CRAN.Rproject.org/package=graph]. [R package version 1.26.0]..Google Scholar
 Gentleman R, Carey V, Bates D, Bolstad B, Dettling M, Dudoit S, Ellis B, Gautier L, Ge Y, Gentry J, Hornik K, Hothorn T, Huber W, Iacus S, Irizarry R, Leisch F, Li C, Maechler M, Rossini A, Sawitzki G, Smith C, Smyth G, Tierney L, Yang J, Zhang J: Bioconductor: open software development for computational biology and bioinformatics. Genome Biol. 2004, 5 (10): R8010.1186/gb2004510r80. [http://genomebiology.com/2004/5/10/R80].PubMed CentralView ArticlePubMedGoogle Scholar
 Csardi G, Nepusz T: The igraph software package for complex network research. InterJournal. 2006, Complex Systems: 1695, [http://igraph.sf.net].Google Scholar
 Müller LAJ, Kugler KG, Dander A, Graber A, Dehmerm M: QuACN  An R package for analyzing complex biological networks quantitatively. Bioinformatics. 2011, 27 (1): 140141. 10.1093/bioinformatics/btq606.View ArticleGoogle Scholar
 Dehmer M, EmmertStreib F, Tsoy Y, Varmuza K: Quantifying structural complexity of graphs: information measures in mathematical chemistry. Quantum Frontiers of Atoms and Molecules. Edited by: Putz M. 2011, Nova Publishing, 479498.Google Scholar
 Guha R: Chemical informatics functionality in R. J Stat Software. 2007, 18 (6): 116.Google Scholar
 Steinbeck C, Han Y, Kuhn S, Horlacher O, Luttmann E, Willighagen E: The Chemistry Development Kit (CDK): an opensource Java Library for Chemo and Bioinformatics. J Chem Inf Comput Sci. 2003, 43 (2): 493500. 10.1021/ci025584y. [http://pubs.acs.org/doi/abs/10.1021/ci025584y]. [PMID: 12653513]..View ArticlePubMedGoogle Scholar
 Cao Y, Charisi A, Cheng CL, Jiang T, Girke T: ChemmineR: a compound mining framework for R. Bioinformatics. 2008, 24 (15): 17331734. 10.1093/bioinformatics/btn307.PubMed CentralView ArticlePubMedGoogle Scholar
 Symyx Solutions Inc:: 2010,, [http://accelrys.com/products/informatics/cheminformatics/ctfileformats/nofee.php]. [CT File Formats, Symyx Solutions, Inc.].
Copyright
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.