QTLNetworkR: an interactive R package for QTL visualization

QTLNetworkR: an interactive R package for QTL visualization 1 first-author 2 1 corresp ‡Corresponding Author ‡Corresponding Author †E-mail:jzhu@zju.edu.cn †E-mail:jzhu@zju.edu.cn jzhu@zju.edu.cn Institute of Bioinformatics, Zhejiang University, Hangzhou 310029, China Institute of Bioinformatics, Zhejiang University, Hangzhou 310029, China Queensland Institute of Medical Research, Brisbane, Queensland 4006, Australia Queensland Institute of Medical Research, Brisbane, Queensland 4006, Australia QTLNetworkR is an R package that aims to provide a user-friendly and platform-independent tool to visualize quantitative trait loci (QTL) mapping results. The graphical functions of the QTLNetworkR are based upon lattice and grid packages, and the graphical user interface (GUI) of the QTLNetworkR is built upon RGtk2 and gWidgetsRGtk2 packages. Six functions are designed to help visualize marker interval, putative QTL, QTL-by-environment interactions, marker interval interactions, epistasis, and the predicted genetic architecture of complex traits. It is especially helpful in profiling results for multiple traits at multiple environments. The current version of QTLNetworkR is able to accept QTL mapping results from QTLNetwork, and it is ready for possible extensions to import results from some other QTL mapping software packages. In addition, we presented a QTL mapping result in rice ( Oryza sativa ) as an example to describe the features of QTLNetworkR. Visualization Quantitative trait loci (QTL) QTL mapping R package QTLNetworkR Introduction 1 Mapping quantitative trait loci (QTL) is of great importance to understand the genetic architecture underlying complex traits (Mackay, 2001 2001 ; Doebley, 2004 2004 ), and many statistical methods have been proposed in the last two decades, such as interval mapping (Lander and Botstein, 1989 1989 ; Haley and Knott, 1992 1992 ), composite interval mapping (Zeng, 1994 1994 ), and multiple interval mapping (Kao et al., 1999 1999 ). Although QTL mapping packages such as MAPMAKER/QTL (Lander et al., 1987 1987 ), QTLMapper (Wang et al., 1999 1999 ), Map Manager (Manly et al., 2001 2001 ), QTL Express (Seaton et al., 2002 2002 ), R/qtl (Broman et al., 2003 2003 ), MCQTL (Jourjon et al., 2005 2005 ), WinQTLCart (Wang et al., 2006 2006 ), and QGene 4.0 (Joehanes and Nelson, 2008 2008 ) have been developed and released to public, none of them can map and visualize epistasis and QTL-by-environment (QE) interactions simultaneously. Recently, a promising tool called QTLNetwork (Yang et al., 2008 2008 ) has been developed based on a novel QTL mapping strategy (Yang et al., 2007 2007 ) that can map single QTLs, epistasis, and QE interaction for experimental populations of various crosses. The command-line versions of QTLNetwork enable it run across most operating systems, such as Microsoft Windows, Mac OS, and Linux, but the graphical user interface (GUI) one is available for Windows only. Despite of its great visualization features, it remains drawbacks in the GUI version; e.g., it is merely able to illustrate one trait at a time and pale to present the QE interaction in detail. Moreover, the profiles from QTLNetwork are insufficient to prepare manuscripts as it lacks flexibility to change some primitives, such as fonts and scales of axis. In order to enhance the quality of the illustration for mapping results, we developed a user-friendly and platform-independent tool called QTLNetworkR to visualize QTL mapping results generated by QTLNetwork in a more convenient and informative way. Six functions were designed in QTLNetworkR for F -statistic profiles of marker interval analysis (looking for marker intervals potentially bound up with QTLs), one-dimensional (1D) genome scan for putative QTLs, QE interaction, and marker interval interaction analyses, two-dimensional (2D) genome scan for epistasis and the predicted genetic architecture. Design and implementation 1 We designed and developed the QTLNetworkR as an add-on package in R (Ihaka and Gentleman, 1996 1996 ) for its popularity in the academic community and platform-independence. To facilitate users, we provided a GUI for the QTLNetworkR based on RGtk2, an add-on R package for combining GUI toolkit GTK+ 2.0 with R, via the aid of another package gWidgetsRGtk2, both of which can be downloaded from the Comprehensive R Archive Network (CRAN, http://cran.r-project.org/ http://cran.r-project.org/ ). There are two alternative graphics systems in R, traditional graphics like S and grid, and the latter one, at least in this study, convinced us that it is a more flexible scaffold with adding further details for QTL mapping, e.g., epistasis and QE interactions. We consequently chose it as the basis of graphical functions in the QTLNetworkR. The QTLNetworkR starts with a ‘file’ container for loading a QTL mapping result file (QTLNetwork output) and a linkage map file. In addition, it is embedded with a dialog notebook accommodating six tabbed pages, for functions designed, with options to control graphical output primitives, e.g., lines and text, to meet the requirement for preparing manuscripts in different formats. Here, we present a QTL mapping result in rice ( Oryza sativa ) as an example to demonstrate the application of QTLNetworkR (unpublished data). It is a recombinant inbred line population on 10 morphological and yield-related traits from seven different environments. A total of 221 RFLP markers scored along the 12 rice chromosomes were used to construct the genetic linkage map and distributed over 15 linkage groups eventually. The QTLNetworkR provides more enhanced visualization than the QTLNetwork (GUI version), including F -statistic profiles of 1D genome scan for the putative QTLs (Fig. 1a 1a ) and 2D genome scan for the epistasis as well as a plot to summarize the network among QTL detected (Fig. 1b 1b ). (a) F -statistic profile of one-dimensional (1D) genome scan on linkage groups (LGs) 1 and 4 for grain rate (red), 1000 grain weight (blue), and full grain number (green), respectively. Linkage peaks that exceed the threshold F -value are considered as putative QTLs; (b) Predicted genetic architecture for grain rate (square), 1000 grain weight (triangle), and full grain number (circle). The loci with black fill represent epistatic QTLs without individual effects. QTLs with potentially pleiotropic effects are labeled by open rectangles. A: additive effect; AE: additive×environment interaction effect; AA: additive×additive epistatic effect; AAE: AA×environment interaction effect Furthermore, pleiotropy often occurs when a single locus influences multiple phenotypic traits, and many lines of evidence espoused its ubiquitousness (Wright, 1980 1980 ; Clark et al., 2006 2006 ). Growing experimental data are collected upon multiple traits and it would be of interest to discover QTLs of pleiotropic effects, if there is any. As the QTLNetwork is mute to profile multiple traits simultaneously, we offered an option to present multiple traits for 1D genome scan and the summarized QTL network to highlight the pleiotropic effects of QTLs (Figs. 1a and 1b 1a and 1b ). In this way, we may find that an epistatic QTL, without individual effect on one trait, could be a main-effect QTL for another trait, e.g., QTL 1-3 (Fig. 1b 1b ). QE interaction is an essential nature for the genetic basis of complex trait. As such information has been limitedly presented in the QTLNetwork, we provided a more detailed illustration to show the main effect (A) across environments and the significant QTL effect (A+AE) in specific environments (Fig. 2 2 ). ‘QTL’ represents the main effect of each putative QTL and ‘Q+QE1’ to ‘Q+QE7’ represent genetic effects (main QTL effects plus QE interaction effects) in 1 to 7 environments. The x -axis label, for example, ‘1–9’ represents a putative QTL locates at marker interval 9 on linkage group 1 The QTLNetworkR package is freely available at http://cran.r-project.org/web/packages/QTLNetworkR/index.html http://cran.r-project.org/web/packages/QTLNetworkR/index.html . Alternatively, it can be downloaded and installed directly in R environment with command: install.packages (pkgs=“QTLNetworkR”). Detailed information about QTLNetworkR can be looked up in its user manual from http://qtlnetworkr.r-forge.r-project.org/ http://qtlnetworkr.r-forge.r-project.org/ Conclusions 1 QTLNetworkR provides a user-friendly and platform-independent tool to visualize QTL mapping results generated but not limited by QTLNetwork. Results from other QTL mapping packages, after reorganizing their formats, can also be imported and visualized. It is capable of presenting QTL mapping results for multiple traits and providing additional options to pronounce QE interaction. It can be easily run in R environment regardless of operating systems. As a QTL mapping visualization tool, the QTLNetworkR will be particularly helpful for profiling results for multiple traits at multiple environments. References KW Broman , H Wu , S Sen , GA Churchill . R/qtl: QTL mapping in experimental crosses . Bioinformatics , 2003 . 19 ( 7 ): 889 - 890 . 12724300 DOI: 10.1093/bioinformatics/btg112 http://doi.org/10.1093/bioinformatics/btg112 . R/qtl: QTL mapping in experimental crosses RM Clark , TN Wagler , P Quijada , J Doebley . A distant upstream enhancer at the maize domestication gene tb1 has pleiotropic effects on plant and inflorescent architecture . Nature Genetics , 2006 . 38 ( 5 ): 594 - 597 . 16642024 DOI: 10.1038/ng1784 http://doi.org/10.1038/ng1784 . A distant upstream enhancer at the maize domestication gene tb1 has pleiotropic effects on plant and inflorescent architecture J Doebley . The genetics of maize evolution . Annual Review of Genetics , 2004 . 38 ( 1 ): 37 - 59 . 15568971 DOI: 10.1146/annurev.genet.38.072902.092425 http://doi.org/10.1146/annurev.genet.38.072902.092425 . The genetics of maize evolution CS Haley , SA Knott . A simple regression method for mapping quantitative trait loci in line crosses using flanking markers . Heredity , 1992 . 69 ( 4 ): 315 - 324 . 16718932 . A simple regression method for mapping quantitative trait loci in line crosses using flanking markers R Ihaka , R Gentleman . R: a language for data analysis and graphics . Journal of Computational and Graphical Statistics , 1996 . 5 ( 3 ): 299 - 314 . DOI: 10.2307/1390807 http://doi.org/10.2307/1390807 . R: a language for data analysis and graphics R Joehanes , JC Nelson . QGene 4.0, an extensible Java QTL-analysis platform . Bioinformatics , 2008 . 24 ( 23 ): 2788 - 2789 . 18940826 DOI: 10.1093/bioinformatics/btn523 http://doi.org/10.1093/bioinformatics/btn523 . QGene 4.0, an extensible Java QTL-analysis platform MF Jourjon , S Jasson , J Marcel , B Ngom , B Mangin . MCQTL: multi-allelic QTL mapping in multi-cross design . Bioinformatics , 2005 . 21 ( 1 ): 128 - 130 . 15319261 DOI: 10.1093/bioinformatics/bth481 http://doi.org/10.1093/bioinformatics/bth481 . MCQTL: multi-allelic QTL mapping in multi-cross design CH Kao , ZB Zeng , RD Teasdale . Multiple interval mapping for quantitative trait loci . Genetics , 1999 . 152 ( 3 ): 1203 - 1216 . 10388834 . Multiple interval mapping for quantitative trait loci ES Lander , D Botstein . Mapping mendelian factors underlying quantitative traits using RFLP linkage maps . Genetics , 1989 . 121 ( 1 ): 185 - 199 . 2563713 . Mapping mendelian factors underlying quantitative traits using RFLP linkage maps ES Lander , P Green , J Abrahamson , A Barlow , MJ Daly , SE Lincoln , LA Newberg . MAPMAKER: an interactive computer package for constructing primary genetic linkage maps of experimental and natural populations . Genomics , 1987 . 1 ( 2 ): 174 - 181 . 3692487 DOI: 10.1016/0888-7543(87)90010-3 http://doi.org/10.1016/0888-7543(87)90010-3 . MAPMAKER: an interactive computer package for constructing primary genetic linkage maps of experimental and natural populations TF Mackay . The genetic architecture of quantitative traits . Annual Review of Genetics , 2001 . 35 ( 1 ): 303 - 339 . 11700286 DOI: 10.1146/annurev.genet.35.102401.090633 http://doi.org/10.1146/annurev.genet.35.102401.090633 . The genetic architecture of quantitative traits KF Manly , RHJr Cudmore , JM Meer . Map Manager QTX, cross-platform software for genetic mapping . Mammalian Genome , 2001 . 12 ( 12 ): 930 - 932 . 11707780 DOI: 10.1007/s00335-001-1016-3 http://doi.org/10.1007/s00335-001-1016-3 . Map Manager QTX, cross-platform software for genetic mapping G Seaton , CS Haley , SA Knott , M Kearsey , PM Visscher . QTL Express: mapping quantitative trait loci in simple and complex pedigrees . Bioinformatics , 2002 . 18 ( 2 ): 339 - 340 . 11847090 DOI: 10.1093/bioinformatics/18.2.339 http://doi.org/10.1093/bioinformatics/18.2.339 . QTL Express: mapping quantitative trait loci in simple and complex pedigrees DL Wang , J Zhu , ZK Li , AH Paterson . Mapping QTLs with epistatic effects and QTL×environment interactions by mixed linear model approaches . Theoretical and Applied Genetics , 1999 . 99 ( 7-8 ): 1255 - 1264 . DOI: 10.1007/s001220051331 http://doi.org/10.1007/s001220051331 . Mapping QTLs with epistatic effects and QTL×environment interactions by mixed linear model approaches SC Wang , CJ Basten , ZB Zeng . Windows QTL Cartographer 2.5 , 2006 . (Available from: http://statgen.ncsu.edu/qtlcart/WQTLCart.htm http://statgen.ncsu.edu/qtlcart/WQTLCart.htm ) . Windows QTL Cartographer 2.5 S Wright . Genic and organismic selection . Evolution , 1980 . 34 ( 5 ): 825 - 843 . DOI: 10.2307/2407990 http://doi.org/10.2307/2407990 . Genic and organismic selection J Yang , J Zhu , RW Williams . Mapping the genetic architecture of complex traits in experimental populations . Bioinformatics , 2007 . 23 ( 12 ): 1527 - 1536 . 17459962 DOI: 10.1093/bioinformatics/btm143 http://doi.org/10.1093/bioinformatics/btm143 . Mapping the genetic architecture of complex traits in experimental populations J Yang , C Hu , H Hu , R Yu , Z Xia , X Ye , J Zhu . QTLNetwork: mapping and visualizing genetic architecture of complex traits in experimental populations . Bioinformatics , 2008 . 24 ( 5 ): 721 - 723 . 18202029 DOI: 10.1093/bioinformatics/btm494 http://doi.org/10.1093/bioinformatics/btm494 . QTLNetwork: mapping and visualizing genetic architecture of complex traits in experimental populations ZB Zeng . Precision mapping of quantitative trait loci . Genetics , 1994 . 136 ( 4 ): 1457 - 1468 . 8013918 . Precision mapping of quantitative trait loci We thank Dr. Guo-bo CHEN (Department of Agronomy, Zhejiang University, China) for his helpful comments. 10.1631/jzus.B0900384 Project supported by the National Basic Research Program (973) of China (No. 2008CB117002), the Ministry of Agriculture Public Benefit Research Foundation of China (No. 200803034), and the National High-Tech R & D Program (863) of China (No. 2006AA10A102) 2009-11-30 2010-05-17 2010-03-09 2022-01-05 Copyright © Zhejiang University and Springer-Verlag Berlin Heidelberg 2010 2010 Journal of Zhejiang University-SCIENCE B (Biomedicine and Biotechnology) 07 11