User Manual∗
Contents
1. Figure 25 BaySeq Interface 34 Likelihood FDR DE id rowID control t1 control 2 treated 1 treated2 ENSG 971 row_7 1 1 1 1 0 261 ENSG 419 row_3 1132 1070 1088 1138 0 217 ENSG 457 row_4 354 348 392 377 0 111 ENSG 003 row 633 590 618 661 0 074 ENSG 460 row_5 618 580 653 621 0 067 ENSG 005 row_2 0 1 0 0 0 051 Figure 26 BaySeq result file Bayseq reports the input counts and the number of the row rowID in the first columns and the Likelihood and the false discovery rate FDR DE in the remaining columns Estimation Type e g quantile the SampleSize e g 1000 an FDR level SampleA e g treated and SampleB e g control The BaySeq function returns two text files and two plots The first text file shows the results of this method see Figure 26 while the second text file shows the differentially expressed genes only The output count file is saved with the name specified by the user in the Name field see Figure 25 If no name is specified by the user then the first file is named with the name of the input file plus _results_BaySeq txt suffix Both text files are saved in the Results folder The second file is named with the name of the input file plus fdr 0 05_DE_genes_BaySeq txt suffix where 0 05 is the chosen FDR for rejection The first plot shows the log ratios of the counts against the mean aver age of the counts and it is named with the name of the input file plus the _Plot2. Figure 35 The demo_ReadCount txt file saved in the Results folder Read Count Interface is ready to work on MyProject project Select a bam folder media 6b8a4404 F3e2 4Fbc 9Fc4 Sa0ebc8d0635 Bam demo 2L_1 bam Open Provide a GTF e2 4fbc 9fc4 5a0ebc8d0635 2L_Drosophila_melanogaster BDGPS 70 gtF Open Counting Mode Union ignore Strand Save outputs Save Results SM Name 2L_counts csv Start Counting Reads Howto use this InterFace CounkGeacs Select a bam Folder by clicking on the corresponding Open button To select the entire bam folder select just one bam file inside the bam folder you want to use The entire Folder will be loaded Provide a GTF file by clicking on the corresponding Open button In Counting Mode field choose one of the 3 modes Union IntersectionStrict and intersectionNotEmpty in Ignore Strand check box you can decide whether the counting procedure must ignore the strand or not Finally click on Count Reads button For Further information see www bioconductor org packages release bioc html GenomicRanges html Figure 36 Read Count Interface 44 id 2L1 2L3 2L4 2L_U1 2LU3 2LU4 2L_U6 FBgn0000018 528 485 546 613 441 501 485 FBgn0000052 2300 2968 3555 2921 3097 3244 2626 FBgn0000053 2361 2982 3790 2307 2352 2542 1856 F Bgn0000055 1 0 0 0 0 0 0 FBgn0000056 0 0 0 0 0 0 0 FBgn0000061 4 2 2 1 1 5 0 FBgn0000075 2 2 1 4 4 3 1 FBgn0000097 3849 3727 4546 4656 4227 344
3. JE E_E fta Bam e1 2L_U6 0 100000 P cq oF o og O 2 ta Bam e1 2L_U1 v oO o P M o 0 100000 8 0 o _d go o o E os o o9 o o 0 80000 6 100000 0 100000 0 100000 0 60000 Figure 41 Count plots generated by the Plot All Counts function 48 PC2 60000 40000 20000 0 20000 40000 80000 PCA Plot 50000 0 50000 PC1 Figure 42 PCA plot generated by the PCA function 49 Heatmap FBgn0010265 FBgn0011272 FBgn0015756 FBgn0020618 FBgn0261606 FBgn0004867 FBgn0261597 FBgn0016075 FBgn0001114 FBgn0261800 FBgn0005593 FBgn0032987 FBgn0003941 FBgn0002593 FBgn0032518 FBgn0031980 FBgn0003942 FBgn0000308 FBgn0015521 FBgn0261608 FBgn0011230 FBgn0043841 FBgn0022893 FBgn0027932 FBgn0027571 FBgn0001942 FBgn0000299 FBgn0261836 FBgn0064225 FBgn0000559 00 2 Ko A l ym 2 cm am e1 2L U3 bam am e1 2L U4 bam Bam e1 2L 1 bam Bam e1 2L 3 bam Bam e1 2L 4 bam qam e1 2L U1 bam Figure 43 Heatmap 50 We click on DESeq button In the DESeq Analysis Interface we select the 2L_counts csv count file We type the T T T U U U U sequence in the Factors field to specify the treated and untreated samples as in EdgeR analysis We type single end paired end paired end single end pair ed end paired end single end in the LibTypes field to specify the libra
4. Now you have to add the bin folder to your PATH variable Make sure you have no other versions of GTK in PATH variable To do this execute the fol lowing instructions Open Control Panel click on System and Security click on System click on Advanced System Settings click on click on Environment Variables In the Environment Variables window you will notice two columns User variables for a user name and System variables Change the PATH vari able in the System variables to be C opt gtk bin 5 Then go to Section 3 3 Installation of R and the required R packages 1 Install R 3 0 2 from http cran r project org according to your operat ing system 2 Download RNASeqGUI package from http bioinfo na iac cnr it RNASeqGUI Download For Windows operating system download the zip binary file For MacOS and Linux download the tar gz file e For Windows users select Install packages s from local zip files under the Packages pull down menu as in the Figure B Eile Edit View Misc Packages Windows Help 21 1x ales 1o R version 2 14 1 2011 Copyright C 2011 The ee aan ISBN 3 900051 07 0 Platform i386 pc mingv Update packages Install package s From local zip files R is free software and comeswroreposerorce tT tre 77700 You are welcome to redistribute it under certain conditions Type license or licence for distribution details PANTY Natural language support
5. Stitt M Usadel B 2012 RobiNA a user friendly integrated software solution for RNASeq based transcriptomics Nucleic Acid Research 40 W1 W622 W627 McCarthy et al 2012 McCarthy D J Chen Y Smyth G K 2012 Differ ential expression analysis of multifactor RNA Seq experiments with respect to biological variation Nucleic Acids Research 40 4288 4297 Morgan et al 2014 Morgan M Carey V Lawrence M 2014 BiocParal lel Bioconductor facilities for parallel evaluation R package version 0 4 1 Mortazavi et al 2008 Mortazavi A Williams B A McCue K Schaef fer L Wold B 2008 Mapping and quantifying mammalian tran scriptomes by RNA seq Nature Methods 5 621 8 Pramana et al 2013 Pramana S 2013 neaGUI An R package to perform the network enrichment analysis NEA R package version 1 0 0 Risso et al 2011 Risso D Schwartz K Sherlock G Dudoit S 2011 GC Content Normalization for RNA Seq Data BMC Bioinformatics 12 1 480 Robinson et al 2010 Robinson M D McCarthy D J Smyth G K 2010 edgeR a Bioconductor package for differential expression analysis of digital gene expression data Bioinformatics 26 139 140 Robinson et al 2007 Robinson M D McCarthy D J Smyth G K 2007 Moderated statistical tests for assessing differences in tag abundance Bioinformatics 23 2881 2887 Robinson et al 2008 Robinson M D McCarthy D J Smyth G K 2008 Small sampl
6. and a text file of normalized counts saved in Results folder Subsequently we use the PCA function by typing the 1 3 4 U1 U3 U4 U6 45 RNASeqGUI Data Exploration interface is ready to work on MyProjects project Select a count file 2L_counts csv Open Headers amp M whatis the Column Separator PCA Factors 1 3 4 U1 U3 U4 U6 Column Column2 log M Howmany genes inthe Heatmap 30 Legend positioninPCA topright How to use this Interface Look at BAM files Plot Pairs of Counts Plot All Counts Count Distr Density MDPlot MeanVarPlot Heatmap S PCA 3 L PCA3D Component Histogram QPlot Histogram QPlot Density Figure 38 Data Exploration Interface sequence in the PCA Factors field see Figure 38 to specify the labels that will be displayed in the legend at the top right of the plot generated by this function shown in Figure 42 Finally we can use the HeatMap function to see what are the first say thirty most expressed genes Therefore we typed the number 30 in the How many genes in the Heatmap field see Figure 43 From the heatmap we can notice that the the most expressed gene is the one called FBgn0000559 look at the bottom of the Figure 1491 Now we can start with the analysis We decide to use EdgeR DESeq and NOISeq and compare the results among them We click on Data Analysis Interface button We start the EdgeR analysis by clicking on the EdgeR button In the EdgeR Analysis Interface we selec
7. is named with the 30 id baseMean baseMeanA baseMeanB foldChange log2FoldChange ENSG 0003 625 025 630 902 619 147 0 981 0 027 ENSG 0005 0 264 0 528 0 0 Inf ENSG 0419 1106 882 1136 118 1077 646 0 948 0 076 ENSG 0457 367 367 362 361 372 374 1 027 0 039 ENSG 0460 617 493 618 055 616 931 0 998 0 002 Figure 20 DESeq output The first column reports the gene ids baseMean reports the mean normalised counts averaged over all samples from both conditions baseMeanA reports the mean normalised counts from condition A baseMeanB mean normalised counts from condition B foldChange re ports the fold changes from condition A to B log2FoldChange reports the logarithm to basis 2 of the fold changes pval reports the p values for the statistical significance and padj reports the p values adjusted for multiple testing calculated by the Benjamini Hochberg algorithm Please choose one of the methods below to identify DE genes You are working on MyProject project EdgeR DESeq DEReq2 RNASeqGUI DESeq2 Interface is ready to work on MyProject project Select a count file Open Headers whatis the Column Separator Factors LibTypes Treated Control Padj 0 05 Save outputs Save Results Name csv How to use this Interface Run DESeq2 Figure 21 DESeq2 interface 31 pval 0 774 0 985 0 297 0 744 0 982 padj 1 1 0 935 1 1 id baseMean log2FoldChange lfcSE stat pvalue ENSG00000000003 625 025 0 025 0 07
8. that are written inside the calculateGUI1 R file Create window window lt gtkWindow At this point MY OWN FUNCTION button is created and the result is the one shown in Figure 47 By clicking this button we call MyHeatmapConn function that calls MyHeatmap function defined before 59 attached base packages 1 grid splines parallel stats graphics grDevices utils 8 datasets methods base other attached packages 1 ineq_0 2 10 e1071_1 6 1 class_7 3 7 4 RGtk2_2 20 27 BiocParallel_0 4 1 scatterplot3d_0 3 35 7 preprocessCore_1 24 0 lLeeBamViews_0 99 24 BSgenome_1 30 0 10 EDASeq_1 8 0 aroma Light_1 32 0 matrixStats_0 8 14 13 ShortRead_1 20 0 VennDiagram_1 6 5 RColorBrewer_1 0 5 16 gplots_2 12 1 DESeq2_1 2 10 RcppArmadillo_0 4 000 4 19 Rcpp_0 11 0 DESeq_ 1 14 0 lattice_0 20 15 22 locfit_1 5 9 1 NOISeq 2 4 0 baySeq_1 16 0 25 edgeR_3 4 2 Limma_3 18 13 pasilla_0 2 19 28 DEXSeq_1 8 0 ggplot2_0 9 3 1 Rsamtools_1 14 3 31 Biostrings_ 2 30 1 GenomicFeatures_1 14 3 AnnotationDbi_1 24 0 34 Biobase_2 22 0 GenomicRanges 1 14 4 XVector_0 2 0 37 IRanges_1 20 6 BiocGenerics_0 8 0 digest_0 6 4 loaded via a namespace and not attached 1 annotate_1 40 0 BatchJobs_1 2 BBmisc_1 5 4 biomaRt_2 18 0 bitops 1 0 6 brew_1 0 6 7 caTools_1 16 codetools_0 2 8 colorspace 1 2 4 10 DBI_0 2 7 dichromat_2 0 0 Fatl_1 2 13 foreach_1 4 1 gdata_2 13 2 genefilter_1 44 0 16 geneplotter_1 40 0 gtable_0 1 2 gtools_ 3 3 0 19 hwrit
9. use this Interface P Venn Diagrams 1 Figure 29 Result Comparison Interface For each function e g FDR Hist P value Hist Likelihood Hist of each method we just need to provide a full result file placed in the Results folder For Volcano Plot and Plot FC functions we must provide a path to a full result file as the one shown in Figure 17 and a FDR P value or Prob value it depends on the chosen method to point out the differentially expressed genes shown in red In this case it is also possible to provide a gene id provided into the Gene Id field to point out that particular gene in the Volcano or FC plot that gene will be displayed in green All generated plots are saved in pdf format in the Plots folder 5 6 2 Result Comparison Interface The second interface includes the possibility to generate Venn diagrams of either two or three result text files See Figure B9 The user must provide two or three text files reporting the results of the used methods and the corresponding labels to recognize these files in the generated diagrams A Venn diagram is generated and saved in the Plots folder Moreover a text file showing the gene ids belonging to the intersection of the selected methods is created and saved in the Results folder 38 5 7 The summary report All the functionalities used by the user are automatically saved in a report file as the one shown in Figure 8 inside the Logs directory of the
10. 000000000000000000000000000000000000000 1000000000000000000000000000000000000000 1 0000000000000000000000000000000000000000 10000000000000000000000000000000000000000 10000000000000000000000000000000000000000 0000000000000000000000000000000000000000 0000000000000000 000000000000000000000000 10000000000000 000000000000000000000000000 1 o000000000000000000000000000000000000000 10000000000000 000000000000000000000000000 0000000000000000000000000000000000000000 0000000000000 000000000000000000000000000 1 0000000000000000000000000000000000000000 1 0000000000000000000000000000000000000000 10000000000000000000000000000000000000000 T T T T T nb ne 02 Al n 23 25 27 29 31 33 35 37 17 19 21 9 11 13 15 Per Base Quality BoxPlot Per Base Quality of Reads of the bam files stored in the folder demo without the 2L_1 bam file Figure 33 Read Count Histogram Read Count Histogram of the bam files stored in the folder demo 1 2e 07 7 1 0e 07 8 0e 06 4 6 0e 06 4 4 0e 06 2 0e 06 4 0 0e 00 90 T2 Le weg ereq vn T2 Le weg ereq en Te Le weg eieq 10 T2 Le wegereq gt Te Le weg ereq Te Le weg ereq T2 Le weg ereq 43 Figure 34 fileName NumberOfReads Data Bam demo 2L_1 12320205 Data Bam demo 2L_3 6477978 Data Bam demo 2L_4 7741241 Data Bam demo 2L_U1 9473462 Data Bam demo 2L_U3 6586330 Data Bam demo 2L_U4 6071744 Data Bam demo 2L_U6 5883666
11. 05_DE_genes_EDGER txt Headers amp Column Separator Third label EDGER Venn Diagrams Venn Diagrams VennDiagram 2 sets DE Vennbiagrayg 3 sets DE Figure 45 Result Comparison Interface Venn Diagram of DE genes NOISEQ DESEQ Figure 46 Venn Diagram 53 7 How to customize RNASeqGUI It is extremely easy to add new buttons that calls new functions Hence a user can customize RNASeqGUI interfaces for his purposes and benefits by adding the methods he needs mostly 7 1 Adding a new button in just three steps For the sake of example suppose you have written a function that generates a heat map as the one written below MyHeatmap lt function x geneNum require RColorBrewer n lt as numeric geneNum x lt as matrix x means rowMeans x select order means decreasing TRUE 1 n show first n genes hmcol colorRampPalette brewer pal 7 Greens 100 heatmap x select col hmcol margins c 5 8 main MyHeatMap If you want to add MyHeatmap function to RNASeqGUI follow these tree simple steps 1 Place MyHeatmap function in a file for instance called MyHeatmap R in the R folder inside the RNASeqGUL_0 0 4 directory 2 Open calculateGUI1 R file This is the file that generates the Data Ez ploration Interface and copy the following 3 lines and paste them at the bottom of this file before parenthesis Here you create the button called MY OWN FUNCTION MYOWNBUTTO
12. 8 2569 Figure 37 The 2L_counts csv file created by Count Reads function and saved in the Results folder this action a text file named 2L_counts csv see Figure 37 is generated and saved in the Results folder A file named counts txt is also generated in case the user forgets to use the Save Results check box at the bottom of the interface The column names in Figure B6 follow the alphabetical order of the bam files placed in the demo folder Now we can explore the obtained count file shown in Figure We click on Data Exploration Interface button Once this interface has ap peared on the screen see Figure 38 we select the 2L_counts csv file First we use the BoxPlot and the Plot All Counts functions by clicking the corresponding buttons see Figure B8 The generated plots are shown in Figure B9 and Figure 4 respectively From Figure 39 we can see that all the count means the black lines in the box plot and all the count distributions are almost aligned Therefore we decide not to normalize the counts since a normalization procedure does not seem to be necessary To better understand whether a normalization procedure is needed we can also use the MDPlot by plotting each sample counts by selecting Column1 and Column fields against all the other sample counts Anyway if we use the full quantile normalization procedure by clicking the Full Quantile button in the Normalization Interface we get the plot show in Figure 40
13. 9 0 318 0 750 ENSG00000000005 0 264 0 014 0 020 0 675 0 499 ENSG00000000419 1106 882 0 072 0 062 1 174 0 240 ENSG00000000457 367 367 0 035 0 095 0 365 0 714 ENSG00000000460 617 493 0 002 0 079 0 033 0 973 Figure 22 DESeq2 output The first column reports the gene ids baseMean reports the base mean over all rows log2FoldChange reports the logarithm to basis 2 of the fold changes 1fcSE reports the standard errors stat reports the Wald statistic pval reports the p values for the statistical signif icance and padj reports the p values adjusted for multiple testing calculated by the Benjamini Hochberg algorithm name of the input file plus _results_DESeq2 txt suffix Both text files are saved in the Results folder The second file is named with the name of the input file plus _padj 0 05_DE_genes_DESeq2 txt suffix where 0 05 is the chosen adjusted p value for rejection The first plot shows the dispersion value for a given mean of normalized counts and it is named with the name of the input file plus the _Dispersion_DESeq2 pdf suffix The second plot shows the dispersion mean value for a given mean of normalized counts and it is named with the name of the input file plus the _Dispersion_Mean_DESeq2 pdf suffix The third plot shows the dispersion local value for a given mean of normalized counts and it is named with the name of the input file plus the _Dispersion_Local_DESeq2 pdf suffix All plots are
14. Consiglio i Nazionale delle Ricerche RNASeqGUI User Manual Francesco Russo and Claudia Angelini CNR IAC Naples March 25 2014 This work was supported by the Italian Flagship InterOmics Project PB P05 and by BMBS COST Action BM1006 to Luisa Contents 3 Installation of R and the required R package 4 Quick start 5 Structure of RNASeqGUI main interface 5 1 How to create a new project or select an existing one 5 2 Bam Exploration Sectio 5 4 1 Data Exploration Interface 5 5 3 DESeq2 5 6 Post Analysis Sectio 5 6 2 Result Comparison Interface 5 7 The summary report 6 2 Usage of RNASeqGU How to customize RNASeqGU 7 1 Adding a new button in just three steps 8 Technical Details 7 5 3 Count Section 2 2 2 2 5 4 Pre Analysis Section 0 2 2 2004 5 9 9 BaySeg 2 eo Sra a oe Ge Hee 5 6 1 Result Inspection Interfacg 2 2 2 6 1 Data Preparation 2 2 2 2 2 22224 12 13 13 17 21 23 23 26 27 27 29 30 32 33 36 36 38 39 40 40 41 54 54 56 Acknowledgement 57 RNASeqGuUI 1 Introduction 1 1 Overview of RNASeqGUI R package This manual describes RNASeqGUI R package that is a graphical user in terface for the identification of differentially expressed genes from RNA Seq experiments R http cran r
15. Core biocLite scatterplot3d biocLite BiocParallel 4 Once the installation is complete please check that all the packages listed above have been installed correctly To see this copy and paste the following list into R 3 0 2 to see whether there are errors coming out library e1071 library ineq library RGtk2 library e1071 library RCurl library digest library ggplot2 library RColorBrewer library VennDiagram library XML 11 library biomaRt library DEXSeq library pasilla library GenomicRanges library GenomicFeatures library Rsamtools library edgeR library baySeq library NOISeq library DESeq library DESeq2 library gplots library EDASeq library leeBamViews library preprocessCore library scatterplot3d library BiocParallel In case an error message is displayed repeat step 3 for the missing packages otherwise go to Section 4 4 Quick start If you have successfully gone through the installation you are ready to use RNASeqGUI as follows 1 Open R 3 0 2 2 Type library RNASeqGUI in the R environment Wait for the package to be loaded 3 Finally type RNASeqGUI After that a dialog window as that one shown in Figure 4 will appear and you can start interacting with the program 12 5 Structure of RNASeqGUI main interface The RNASeqGUI main interface is divided into five Sections as shown in Figure 4 Each section correspon
16. ION Data Explo pation Interface Normalization Interface RNASeqGUI Data Exploration Interface is ready to work on Myproject project Select a count file Open Headers whatis the Column Separator PCA Factors Column1 1 Column2 2 log How many genes inthe Heatmap 30 Legend positioninPCA topright How to use this Interface Look at BAM files Plot Pairs of Counts Plot All Counts Count Distr Density Sa MDPlot MeanvarPlot Heatmap PCA PCA3D Component Histogram QPlot Histogram QPlot Density Figure 13 Data Exploration Interface 5 4 Pre Analysis Section The third section of the GUI contains two interfaces Data Exploration In terface see Figure 13 and Normalization Interface see Figure I4 Both interfaces take an input count file that must be tab delimited and must have the structure shown in Figure The rows represent genes ids and the columns represent the samples 5 4 1 Data Exploration Interface In Data Exploration Interface there are twelve methods Plot Pairs of Counts Plot all Counts Count Distr Density MDPlot Mean VarPlot Heatmap PCA PCA3D Component Histogram Qplot Histogram Qplot Density e The Plot Pairs of Counts makes use of plot function of the graphics 23 package This function takes a count file as input in tzt or cvs format where the rows correspond to the gene ids and the columns correspond to the samples This function also takes two inte
17. MA_BaySeqNB pdf suffix The second plot shows the posterior likelihood This plot is named with the name of the input file plus the Posteriors_BaySeqNB pdf suffix This method is very time consuming 35 0 738 0 760 0 803 0 833 0 853 0 869 Data Exploration Interface Normalization Interface DATA ANALYSIS SECTION Data Analysis Interface POST ANALYSIS SECTION Result inspection Interface Result Comparison interface RNASeqGUI Result inspection Interface is ready to work on Myproject project Select a result file Open Headers amp Column Separator Gene Id FDR 0 05 Padj 0 05 Prob 0 7 How to use this Interface EdgeR BaySeq DESeq DESeq2 NoiSeq Figure 27 Result Inspection Interface 5 6 Post Analysis Section In the fifth section of the GUI called Post Analysis Interface there are two interfaces Result Inspection Interface sce Figure 27 and Result Comparison Interface see Figure 29 The first interface includes the possibility to generate several plots for each methods The second allows to compare the outcomes obtained from several methods 5 6 1 Result Inspection Interface To explore the results of a specific method we have to click on the used method in Data Analysis Section say EdgeR and the interface in Figure will display the functions available for the selected method for EdgeR Plot FC FDR Hist P value Hist functions are available If we click all buttons in Figure
18. N lt gtkButtonNewWithMnemonic MY OWN FUNCTION show TRUE Associate the button to MyHeatmapConn that calls MyHeatmap function gSignalConnect MYOWNBUTTON clicked MyHeatmapConn the buttons packStart MYOWNBUTTON f i11 F 3 Finally Copy the following code MyHeatmapConn lt function button user data res lt NULL Get the information about data and the file the file lt filename getText the sep lt sepEntry getText the headers lt headersEntry active the geneNum lt geneNum getText d lt read table the file sep the sep header the headers row names 1 Select numerical variables numVar lt sapply 1 ncol d function x is numeric d x if sum numVar 0 error lt ERROR No numerical variables in the data else res MyHeatmap d the geneNum HERE YOU CALL THE FUNCTION YOU DEFINED 54 RNASeqGUI Data Exploration Interface is ready to work on MyProject project Select a count file l Open Headers whatis the Column Separator PCA Factors Column1 1 Column2 2 log SM How many genes inthe Heatmap 30 Legend position in PCA topright Look at count files before the Analysis Plot Pairs of Counts Plot AllCounts Count Distr Density MDPlot MeanvarPlot Heatmap PCA PCA3D Component Histogram QPlot Histogram QPlot Density Figure 47 A new button called MY OWN FUNCTION is created and paste it before the two following lines below
19. OST ANALYSIS SECTION Result Inspection interface Result Comparison Interface INFO Contacts and INFO Licences Terms and Conditions Figure 5 Creation of a new project 15 Please EITHER create a new project OR select an existing one Then choose one of the Interfaces below Choose a Project Name Create a New Project Otherwise choose an existing project MyProject Select Ns project PRE ANALYSIS SECTION Data Exploration Interface NormalizationInterface DATA ANALYSIS SECTION Data Analysis interface POST ANALYSIS SECTION Result Inspection interface Result Comparison Interface INFO Contacts and INFO Licences Terms and Conditions Figure 6 Selection of an existing project Name v Size Type gt jj Logs Titem folder gt Plots Oitems folder gt Results Oitems folder Figure 7 Structure of the MyProject directory 16 the Create a New Project button RNASeqGUI checks whether the RNASeqGUI_Projects folder already exists in your working direc tory If this folder was already created then RNASeqGUI does not create a copy of it and all the projects you will create will be stored in it Now inside RNASeqGUI_Projects you find MyProjects direc tory Inside this directory three folders are automatically created see Figure 7 such as Logs Results Plots In the Logs folder a report txt file is created t
20. SG 6025 0 577 8699 2 8e14 7 le 12 ENSG 6047 0 627 6 027 0 001 0 027 ENSG 6118 0 152 10 456 0 001 0 039 ENSG 6282 0 418 9 966 1 0e 14 3 3e 12 Figure 18 The EdgeR second text file showing the differentially expressed genes only Columns are the same as in Figure If no name is specified by the user then the first output count file is named with the name of the input file plus _results_EdgeR txt suf fix The second file is named with the name of the input file plus fdr 0 05_DE_genes_EdgeR txt suffix where 0 05 is the chosen FDR Both text files are saved in the Results folder The first plot shows the Biological Coefficient of Variation for a given CPM Count Per Million and is named with the name of the input file plus _Dispersion_EdgeR pdf suffix The second plot shows the relative similarities of the samples and is named with the name of the input file plus _MDS_EdgeR pdf suffix Both plots are saved in the Plots folder 5 5 2 DESeq e The DESeq method Anders et al 2010 see Figure 19 takes an in put count file as the one shown in Figure l2 via the Open button and returns two text files and a plot The first text file shows the results of this method see Figure 2Q while the second text file shows the differentially expressed genes only The output count file is saved with the name specified by the user in the Name field see Figure L9 If no name is specified by the user th
21. ation in several different sections and interfaces and via the inclu sions of numerous concise and clear vignettes and also to facilitate the exten sibility of the GUI via its software development organization that facilitate the task of expanding and redesign its interfaces In fact it is extremely easy to add new buttons that calls new functionalities Therefore a user can customize RNASeqGUI interfaces for his own purposes and benefits by adding the methods he needs mostly for more details see Section 7 How to customize RNASeqGUI Adding a new button in just three steps Hence we think that RNASeqGUI represents a useful and valid alternative to other existing GUIs 1 3 Scope and availability RNASeqGUI is an R package designed for the identification of differentially expressed genes across multiple biological conditions This software is not just a collection of some known methods and functions but it is designed to guide the user during the entire analysis process Moreover the GUI is also helpful for those who are expert R users since it speeds up the usage of the included RNA Seq methods drastically Current implementation allows to handle the simple experimental design where the interest is on the exper imental condition future work will cover complex designs RNASeqGUI is freely available at see Figure I http bioinfo na iac cnr it RNASeqGUI Download 2 Installation guide RNASeqGUI package requires the RGTK2 graphi
22. but running in an English locale R is a collaborative project with many contributors Type contributors for more information and citation on how to cite R or R packages in publications Type demo for some demos help for on line help or help start for an HTML browser interface to help Type qQ to quit R gt T eft Figure 2 Select Install packages s from local zip files under the Packages pull down menu From http outmodedbonsai sourceforge net InstallingLocalRPackages html P g amp g e For MacOS users under Package and Data pull down menu select Package Installer In the Package Installer pull down the top left menu select Local Binary Package and navigate to where you have downloaded the binary package as in the Figure 3 e For Linux users open a shell and go to the directory containing the package tree and type the command sudo R CMD INSTALL 1 path to library RNASeqGUI_0 0 4 3 Finally if the libraries required by RNASeqGUI are not automatically downloaded and installed we suggest the user to install all the packages that are needed to run RNASeqGUI package before loading it Open R 3 0 2 and type the order of the list below is important install packages e1071 install packages ineq install packages RGtk2 install packages RCurl Under Package and Data pull down menu select Package Installer R File Edit Fo
23. cal library Lawrence et al 2010 to run The installation process consists in two steps The first depends on the operating system devoted to installation the GTK 2 0 an open source GUI tool written in C The second regards the required R packages 2 1 For Linux users We tested the RNASeqGUI on Ubuntu 12 10 1 Open a terminal and type sudo apt get update sudo apt get install libgtk2 0 dev 2 Type sudo apt get install libcurl4 gnutls dev 3 Type sudo apt get install libxml2 dev 4 Then go to Section 3 RNASeqGUI Home Example Manual Download Contact Material Credits A GUI for the identification of differentially expressed genes Authors Dr Francesco Russo and Dr Claudia Angelini IAC CNR Links RNASeqGUI is a graphical user interface for the identification of differentially expressed genes CNR from RNA Seg experiments RNASeqGUI was implemented following and expanding the idea presented in tuxette chix RNASeqGUI includes several well known RNA Seg tools available as command line in bioconductor org RNASeqGUI is divided into five main sections Each section is dedicated to a particular step of the data analysis process The first section covers the exploration of the bam files The second concerns the counting process of the mapped reads against a genes annotation file The third focuses on the exploration of count data and on preprocessing of the data including the normalization procedures The fourth i
24. d see Figure 25 takes an input count file as the one shown in Figure via the Open but ton a list of factors e g treated treated control control in the Factors field a NDE list e g 1 1 1 1 a DE list e g 1 1 2 2 an 33 id control_mean treated_mean M D prob ranking ENSG00000000003 575 05 582 71 0 019 7 659 0 104 7 659 ENSG00000000005 0 22 0 47 1 083 0 251 0 037 1 112 ENSG00000000419 1000 84 1049 17 0 068 48 333 0 405 48 333 ENSG00000000457 345 75 334 47 0 047 11 275 0 164 11 275 ENSG00000000460 572 81 570 80 0 005 2 004 0 028 2 004 Figure 24 NoiSeq result file The first column reports the gene ids control_mean is the mean across the control samples treated_mean is the mean across the treated samples M is the log2 ratio of the means of the two conditions and D is the difference between the two conditions means prob is the probability of differential expression the ranking is a summary statistic of M and D values equal to sign M x VM D Please choose one of the methods below to identify DE genes You are working on MyProject project EdgeR DESeq DESeq2 NoiSeq Baygeq RNASeqGUI BaySeq InterFace is ready to work on MyProject project Select a count file Open Headers Whatis the Column Separator Factors NDE DE Estimation Type quantile Sample Size 1000 FDR 0 05 SamplesA samplesB Save outputs Save Results _ Name CSV How to use this Interface Run BaySeq with NB
25. directory with the following R commands respectively getwd setwd path you want to set The creation of RNASeqGUI_Projects directory will only occur the first time you start using RNASeqGUI Subsequently when you click 13 RNASeqGuUI Please EITHER create a new project OR select an existing one Then choose one of the Interfaces below Choose a Project Name Create a New Project Otherwise choose an existing project Select this project BAM EXPLORATION SECTION Bam Exploration Interface COUNT SECTION Read Count Interface PRE ANALYSIS SECTION Data Exploration Interface Normalization Interface DATA ANALYSIS SECTION Data Analysis Interface POST ANALYSIS SECTION Result inspection interface Result Comparison interface Contacts and INFO Licences Terms and Conditions Figure 4 Sections of RNASeqGUI main interface 14 Please EITHER create a new project OR select an existing one Then choose one of the Interfaces below Choose a Project Name MyProject Createa Naw Project Otherwise choose an existing project Select this project MyProject project created You have created your MyProject project in the RNASeqGUIProjects folder OK PRE ANALYSIS SECTION Data Exploration Interface NormalizationInterface DATA ANALYSIS SECTION Data Analysis interface P
26. ds to a particular step of the RNA Seq data analysis work flow Each section contains one or more Graphical Interfaces that can be called by clinking the corresponding button Inside each interface there is a How to use this interface button that displays a vignette to help the user to use the interface see Figure and there are several available functionalities also called functions or methods in the rest of the manual Each function takes specific inputs that can be numeric ones strings or both and generate an output that can be a plot a text file or both The sections of RNASeqGUI will be described one by one in the next sections of this manual 5 1 How to create a new project or select an existing one To start using RNASeqGUI you must either create a new project by choos ing a name for it suppose you choose as name MyProject and then clicking on the Create a New Project button see Figure 5 or select an existing project by typing the name and then clicking on the Select this Project button see Figure 6 The two cases are explained below 1 In the first case if you are using RNASeqGUI for the first time a direc tory called RNASeqGUI_Projects is created in your current working directory type getwd in the R environment to know where you are Inside RNASeqGUI_Projects directory a project folder is created with the name chosen by you in this case with the name MyProject At any moment you can see or change your working
27. e we can call it demoProject and then we click on Bam Exploration Interface button We select the demo folder with the Open button After that we start the analysis by us ing the Read Counts button in the Bam Exploration Interface This action creates the plot shown in Figure 34 The bam files in the demo folder are 41 Per Base Mean Quality Plot Data Bam demo 2L_3 bam Data Bam demo 2L_4 bam Data Bam demo 2L_U1 bam Data Bam demo 2L_U3 bam Data Bam demo 2L_U4 bam Data Bam demo 2L_U6 bam Quality 25 30 35 fi W BH 20 i 15 fi 24 th T T T T 0 20 40 60 Cycle Figure 32 Mean Quality of Reads of the bam files stored in the folder demo without the 2L_1 bam file loaded in alphabetically order and their name are displayed at x axis in Fig ure 34 alphabetically This plot is automatically saved in pdf format in the Plots folder of the project you selected A text file is also generated and saved in the Results folder with the demo_Read Count txt name as shown in Figure This file shows the number of reads for each bam file Critical We cannot use the Mean Quality of Reads or Per Base Quality of Reads function for this dataset since the 2L_1 bam file was generated by pulling fastq files containing reads of different length This file correspond to CG8144_RNAi 1 at page 1774 of Anders et al 2013 To use these functions we need bam files containing reads of the same length O
28. e estimation of negative binomial dispersion with appli cations to SAGE data Biostatistics 9 321 332 Robinson et al 2010 Robinson M D Oshlack A 2010 A scaling normal ization method for differential expression analysis of RNA seq data Genome Biology 11 R25 Sanges et al 2007 Sanges R Cordero F Calogero R A 2007 oneChan nelGUI a graphical interface to Bioconductor tools designed for life scientists who are not familiar with R language Bioinformatics 23 3406 3408 59 Smyth et al 2005 Smyth G K 2005 Limma linear models for microar ray data Bioinformatics and Computational Biology Solutions using R and Bioconductor Springer 397 420 Soneson et al 2013 Soneson C Delorenzi M 2013 A comparison of methods for differential expression analysis of RNA seq data BMC Bioinformatics 14 e91 Tarazona et al 2011 Tarazona S Garcia Alcalde F Ferrer A Dopazo J Conesa A 2011 Differential expression in RNA seq a matter of depth Genome Research 21 2213 222 Villa Vialaneix et al 2013 Villa Vialaneix N Leroux D 2013 sexy rgtk a package for programming RGtk2 GUI in a user friendly manner In Proceedings of 2mes rencontres R Wettenhall et al 2006 Wettenhall J M Simpson K M Satterley K Smyth G K 2006 affylmGUI a graphical user interface for lin ear modeling of single channel microarray data Bioinformatics 22 897 899 Wettenhall et al 2004 Wettenhal
29. e fields as shown in Figure 45 We click on VennDiagrams3setsDE button This action creates two files The first file is the pdf shown in Figure 46 and saved in Plots folder The second file is a text file called NOISEQ_DESEQ_EDGER_genes_in_intersection txt and saved in the Results folder This text file reports the 86 gene ids that fall in the intersection of all the three methods see in Figure 46 All the functionalities we have used are automatically saved in a report file inside the Logs directory 51 EdgeR Fold Change Plot RNASeqGUI co Exploration of the Results Select a file 0 0 0 Results 2L_counts csv_results_EdgeR txt Headers Column Separator Gene Id FBgn0000559 FDR 0 05 Padj 0 05 Prob 0 9 EdgeR BaySeq DESeq ae Gare A teaser A DESeq2 NoiSeq t EdgeR Analysis a r Plot RC FDRHist Volcano Plot N 4 e e a R z O LL D FBgn08005 s N l 0 5 10 15 logqCPM Figure 44 Fold Change Plot generated by using the function PlotFC of EdgeR 52 RNASeqGUI COMPARE RESULTS Select the First file to Compare NASeqGUI_0 0 0 Results 2L_counts csv_prob 0 6_DE_genes_NOISeq txt Headers MM ColumnSeparator First label NOISEQ Select the Second file to Compare NASeqGUI_0 0 0 Results 2L_counts csv_padj 0 05 DE_genes_DESEQ txt Headers amp Column Separator Second label DESEQ 7 Select the Third file to Compare RNASeqGUI_0 0 0 Results 2L_counts csv_fdr 0
30. e of plotQuality function of the EDASeq package Risso et al 2011 This function returns a plot showing the quality of each base of the reads averaged across all bam files e The Per Base Quality of Reads makes use of plotQuality function of the EDASeq package Risso et al 2011 This function returns as many box plots as the number of bam files stored in the provided input folder Each box plot shows the quality of the reads per each base This function makes use of bplapply function of the BiocParallel package Morgan et al 2014 to parallelize the code in order to reduce the execution time e The Reads Per Chromosome makes use of barplot function of the graphics package This function returns as many histograms as the number of bam files stored in the provided input folder Each histogram shows the number of reads are present in each chromosome 18 Please EITHER create a new project OR select an existing one Then choose one of the Interfaces below Choose a Project Name Create a New Project Otherwise choose an existing project MyProject Select this project BAM EXPLORATION SECTION Bam Exploration Interface RNASeqGUI Bam Exploration Interface is ready to work on MyProject project Select a bam folder How to use this Interface Look at BAM files Read Counts Mean Quality of Reads Per Base Quality of Reads Reads Per Chromosome Nucleotide Fr
31. em of log 0 The Qplot Density makes use of qplot function of the ggplot2 pack age This function takes a count file and and returns a plot showing the density function of each column in the count file Moreover for this 25 Bam Exploration Interface COUNT SECTION Read Count Interface PRE ANALYSIS SECTION Data Exploration Interface Normalization Interface RNASeqGUI Normalization Interface is ready to work on Myproject project Select a count file Open Headers Whatis the Column Separator How to use this Interface Normalization Procedures RPKM Upper Quartile TMM FullQuantile Figure 14 Normalization Interface function it is possible to generate the density either of the raw counts or the log of the counts we add 1 to each number in the count file to avoid the problem of log 0 5 4 2 Normalization Interface The Normalization Interface see Figure 4 includes four normalization procedures RPKM Upper Quartile TMM Full Quantile e RPKM makes use of rpkm function of the NOISeq package Tarazona et al 2011 This function takes a count file as specified in Figure 12 and returns a count file with normalized numbers This function performs the RPKM Mortazavi et al 2008 normalization e Upper Quartile makes use of uqua function of the NOISeq package Tarazona et al 2011 This function takes a count file as specified in Figure 2 and returns a count file with normalized number
32. en the first output count file is named with the name of the input file plus _results_DESeq txt suffix 29 Please choose one of the methods below to identify DE genes You are working on MyProject project Ed geR 06569 DESeq2 NoiSeq BaySeq RNASeqGUI DESeq Interface is ready to work on MyProject project Select a count file Open Headers wWhatis the Column Separator Factors LibTypes Treated Control Padj 0 05 Save outputs Save Results Name CSV How to use this Interface Run DESeq Figure 19 DESeq interface The second file is named with the name of the input file plus _padj 0 05_DE_genes DESeq txt suffix where 0 05 is the chosen p value adjusted Both text files are saved in the Results folder The generated plot shows the dispersion value for a given mean of normalized counts This plot is named with the name of the input file plus Dispersion DESeq pdf suffix and it is saved in the Plots folder 5 5 3 DESeq2 e The DESeq2 method Anders et al 2010 see Figure takes an input count file as the one shown in Figure via the Open button and returns two text files and three plots The first text file shows the results of this method see Figure 0 while the second text file shows the differentially expressed genes only The output count file is saved with the name specified by the user in the Name field see Figure 21 If no name is specified by the user then the first file
33. equencies INFO Contacts and INFO Licences Terms and Conditions Figure 9 By clicking the Bam Exploration Interface button in the red cycle the interface to explore bam files will be displayed 19 Bam Exploration Interface is ready to work on MyProject project Selecta bam folder Open Look at BAM files Read Counts MeanQualityof Reads Per Base Quality of Reads Reads Per Chromosome Nucleotide Frequencies Select a bam folder by clicking on the corresponding Open button To select the entire bam folder select just one bam file inside the bam folder you want to use The entire folder will be loaded Then click on the button you want to use For further information see http bioinfo na iac cnr it RNASeqGUI manual pdF Figure 10 By clicking How to use this Interface button a vignette win dow will appear on the screen 20 This function makes use of bplapply function of the BiocParallel package Morgan et al 2014 to parallelize the code in order to reduce the execution time e The Nucleotide Frequencies makes use of plotNtFrequency func tion of EDASeq package Risso et al 2011 This function returns a plot showing the percentage of each nucleotide at each position of the reads Figures will be stored in folder Plots tables in folder Results 5 3 Count Section In the second section of the GUI you find the Count Reads function al
34. er_1 3 iterators _1 0 6 KernSmooth_2 23 10 22 lLabeling_0 2 LatticeExtra_0 6 26 MASS 7 3 26 25 munsell_0 4 2 plyr_1 8 proto_0 3 10 28 RCurl_1 95 4 1 reshape2_1 2 2 R methodsS3_1 6 1 31 R o0_1 17 0 RSQLite_0 11 4 rtracklayer_1 22 3 34 scales_0 2 3 sendmailR_1 1 2 statmod_1 4 18 37 stats4_3 0 1 stringr_0 6 2 survival_2 37 7 40 tools 3 0 1 XML_3 98 1 1 xtable_1 7 1 43 zlibbioc_1 8 0 Figure 48 Session Info 8 Technical Details To see the versions of the used methods we type sessionInfo and we get the list shown in Figure 56 Acknowledgement We want to thank M Franzese V Costa and R Esposito for suggestions and discussions D Granata for technical support This work was supported by the Italian Flagship InterOmics Project PB P05 and by BMBS COST Action BM1006 57 References Anders et al 2010 Anders S Huber W 2010 Differential expression analysis for sequence count data Genome Biology 11 R106 Anders et al 2013 Anders S McCarthy D J Chen Y Okoniewski M Smyth G K Huber W and Robinson M D 2013 Count based dif ferential expression analysis of RNA sequencing data using R and Bioconductor Nature Protocols 8 1765 1786 Angelini et al 2008 Angelini C Cutillo L De Canditiis D Mutarelli M Pensky M 2008 BATS a Bayesian user friendly software for analyz ing time series microarray experiments BMC Bioinformatics 9 415 Bolstad et al 2003 Bolstad B M Iri
35. f seven samples Three samples represent the response to a treatment and four samples are controls Each sample is a cell cul ture of Drosophila melanogaster For more details about this experiment see 40 BamFileName NameOfTheReducedBam LibraryType LibraryLayout CG8144_RNA 1 2L_1 treated single CG8144_RNA 3 21 3 treated paired CG8144_RNA 4 2L_4 treated paired Untreated 1 21 01 untreated single Untreated 3 2L_U3 untreated paired Untreated 4 2L_U4 untreated paired Untreated 6 2L_U6 untreated single Figure 31 Experimental design Brooks et al 2011 We downloaded and aligned the fastg files by running tophat2 as described in at page 1774 Once the bam files were obtained we called them CG8144_RNA 1 CG8144_RNA 3 CG8144_RNA 4 Untreated 1 Untreated 3 Untreated 4 Untreated 6 as in in Anders et al 2013 it is possible to perform the analysis with RNASeqGUI For illustrative purpose and for keeping the computational cost of the demonstrative example under control we limit our attention to chromosome 2L Alignment data bam files are contained in the folder called demo inside the Bam folder with the following names 2L_1 bam 2L_3 bam 2L_4 bam 2L_U1 bam 2L_U3 bam 2L_U4 bam 2L_U6 bam see Figure 31 6 2 Usage of RNASeqGUI We open R 3 0 2 then we type library RNASeqGUI and we type RNASeqGUI Once the main RNASeqGUI interface see Figure 4 has appeared on the screen we create a new project for instanc
36. gers one specifying Columni and the other specifying Column of the count file see Figure and plots the counts of sample in Column1 against the counts of sample in Column2 Moreover for this function it is possible to plot either the raw counts or the log of the counts we add 1 to each number in the count file to avoid the problem of log 0 The Plot all Counts makes use of plot function of the graphics package This function takes a count file as input and produces all possible plots that can be generated by each column in the file against all the other columns If the input text file has n columns then n n 1 plots will be produced An example of this plot is shown in Figure 41 For this function the log check box does not change anything The Count Distr makes use of boxplot function of the graphics package This function takes a count file as input and generates a box plot showing the distribution of the counts for each column in the file An example of this plot is shown in Figure B9 Moreover for this function it is possible to generate the box plot either of the raw counts or the log of the counts we add 1 to each number in the count file to avoid the problem of log 0 The Density makes use of density function of the stats package This function takes a count file and a sample specified by an integer in Columni as input and produces a curve representing the density function of the counts for the selected sample The method i
37. ibing the interface usage briefly as shown in Figure e The Read Counts makes use of barplot function of the graphics package This function returns an histogram as the one shown in Fig 17 This is the MyProject project report Project created the DATE R version 3 0 1 2013 05 16 Platform x86_64 unknown Linux gnu 64 bit locale 1 LC_CTYPE en_US UTF 8 LC_NUMERIC C 3 LC_TIME en_US UTF 8 LC_COLLATE en_US UTF 8 5 LC_MONETARY en_US UTF 8 LC_MESSAGES en_US UTF 8 7 LC_PAPER C LC_NAME C 9 LC_ADDRESS C LC_TELEPHONE C 11 LC_MEASUREMENT en_US UTF 8 LC_IDENTIFICATION C attached base packages 1 grid splines parallel stats graphics grDevices utils 8 datasets methods base other attached packages 1 ineq_0 2 10 e1071_1 6 1 class_7 3 7 4 RGtk2_2 20 27 BiocParallel_0 4 1 scatterpLlot3d_0 3 3 7 preprocessCore_1 24 0 leeBamViews_0 99 24 BSgenome_1 30 9 10 EDASeq_1 8 0 aroma light_1 32 0 matrixStats_0 8 14 13 ShortRead_1 20 0 VennDiagram_1 6 5 RColorBrewer_1 0 5 16 gplots_2 12 1 DESeq2_1 2 10 RcppArmadillo_0 4 0 1191 Reop 0 11 0 0 560 1 14 0 lattice 0 20 15 Figure 8 An example of the file report txt automatically created in Logs directory at the creation of MyProject project Note that the session information is included ure 34 showing the number of mapped reads in each bam file stored in the input folder and a txt tab delimited file summarizing the counts e The Mean Quality of the Reads makes us
38. id the problem of log 0 The PCA makes use of prcomp function of the stats package This function takes a count file a comma separated sequence of strings e g a b c d indicating what are the labels for the legend to be specified in the field Factors see Figure 3 and Legend position in PCA that can be topright bottomright topleft bottomleft The PCA function returns the principal component analysis plot between the first two components An example of PCA plot is shown in Figure For this function the log check box does not change anything The PCA3D makes use of scatterplot3d function of the scatterplot3d package This function takes the same inputs of the PCA function and returns the 3D PCA plot between the first the second and the third principal component For this function the log check box does not change anything The Component Histogram makes use of screeplot function of the stats package This function takes a count file and returns an his togram showing the variance level of each component For this function the log check box does not change anything The Qplot Histogram makes use of qplot function of the ggplot2 package This function takes a count file and and returns an histogram showing the count level of each column in the count file Moreover for this function it is possible to generate the histogram either of the raw counts or the log of the counts we add 1 to each number in the count file to avoid the probl
39. ity that takes four inputs see Figure I The first input must be the name of the folder containing the bam files we want to process The second input must be an annotation file in GTF format General Transfer For mat The third input specifies the count mode that can be one of the following Union IntersectionStrict and IntersectionNotEmpty The fourth input is Ignore Strand check box that allows to perform a strand specific counting task or not The Count Reads button calls the function summarizeOverlaps from the package GenomicRanges Lawrence et al 2013 to obtain gene counts and returns a data frame as the one shown in Figure The first column of this data frame represents the Gene Id while the other columns correspond to the names of the loaded bam files The other en tries report the number of reads that have hit a particular gene for each sam ple see www bioconductor org packages release bioc vignettes GenomicRanges inst doc summarizeOverlaps pdf for more information about the count ing modes Read counting can be a very computational demanding task especially for large experiments with several samples and big alignment files The R en vironment is not optimized from this point of view Therefore the count ing task can be problematic on standard PC with limited clock speed and memory space In this case it could be beneficial either to process sam ples independently or to import count tables in the format specified in Figu
40. kes an input count file as the one shown in Figure 2 via the Open button and returns two text files and two plots The first text file shows the overall result obtained by edgeR see Fig ure 17 while the second text file extracts the subset of differentially expressed genes only see Figure 18 The output count file is saved with the name specified by the user in the Name field see Figure L6 27 Please choose one of the methods below to identify DE genes You are working on MyProject project 4068 DESeq DESeq2 NoiSeq BaySeq RNASeqGUI EdgeR Interface is ready to work on MyProject project Select a count file Open Headers Whatis the Column Separator Factors FDR 0 05 Save outputs Save Results Name CSV How to use this Interface Run EdgeR Figure 16 EdgeR interface id logFC logCPM PValue FDR ENSG 003 0 023 9 181 0 736 1 ENSG 005 2 357 1 058 1 1 ENSG 938 2 5e 15 0 888 1 1 ENSG 971 0 078 1 472 1 1 Figure 17 The first text file produced by the EdgeR method The first column reports the gene ids LogFC reports the log of the fold changes LogCPM reports the the log of the counts per million PValue reports the p values and FDR reports the false discovery rates calculated by the Benjamini and Hochberg s algorithm 28 id logFC logCPM PValue FDR ENSG 3756 0 151 10 652 0 001 0 035 ENSG 4777 0 523 8 455 2 6e10 4 3e 08 ENSG 5961 0 506 6 340 0 002 0 049 EN
41. l J M Smyth G K 2004 limmaGUI a graphical user interface for linear modeling of microarray data Bioin formatics 20 3705 3706 60
42. o report all the ac tions you perform and which parameters you use by performing those actions A session information that summaries all the versions of the used packages is automatically written in the report txt file see Fig ure B at the creation of the project and each time you star this project again 2 In the second case an existing project is selected see Figure 6 RNASe qGUI checks whether the selected name already exists in the RNASe qGUI _Projects folder If no project with the chosen name is found a message warns the user that the selected project does not exist When an existing project is restarted RNASeqGUI continues to write in the same report txt file created previously 5 2 Bam Exploration Section In the first section of the GUI we find the Bam Exploration Interface see Figure D that can be easily called by clicking the corresponding button In this interface we find five different methods to explore the bam files Read Counts Mean Quality of the Reads Per Base Quality of Reads Reads Per Chromosome Nucleotide Frequencies Each of these func tions takes a folder name as input This input folder must contain all the bam files that the user wants to explore To select the entire bam folder select just one bam file inside the bam folder you want to use The entire folder will be loaded To use this interface you can also click on How to use this Interface button and a vignette window will appear on the screen descr
43. project org is an open source object oriented lan guage for statistical computing and graphics RNASeqGUI package includes several well known RNA Seq tools available as command line in www bioconductor org RNASeqGUI main interface is divided into five sections Each section is dedi cated to a particular step of the data analysis process The first section covers the exploration of the bam files The second concerns the counting process of the mapped reads against a gene annotation file GTF The third focuses on the exploration of count data and on data preprocessing including the normalization procedures The fourth is about the identification of the differ entially expressed genes that can be performed by several methods such as DESeq DESeq2 EdgeR NOISeq BaySeq Finally the fifth section regards the inspection of the results produced by these methods and the quantitative comparison among them 1 2 Other GUIs for RNASeq data analysis This package was implemented following and expanding the idea presented in Villa Vialaneix et al 2013 and in http tuxette nathalievilla org p 866 amp lang en The idea of RNASeqGUI is similar to that one presented in Wettenhall et al 2004 Sanges et al 2007 Lohse et al 2012 Pramana et al 2013 Wettenhall et al 2006 Angelini et al 2008 with specific attention on RNA Seq data analysis More over RNASeqGUI is designed to facilitate RNA seq work flow analysis via its organiz
44. re in RNASeqGUI obtained from other tools such as HTSeq count www huber embl de users anders HTSeq Therefore this function makes use of bplapply function of the BiocParallel package Morgan et al 2014 to parallelize the code in order to reduce the execution time 21 Please EITHER create a new project OR select an existing one Then choose one of the interfaces below Choose a Project Name Create a New Project Otherwise choose an existing project Myproject Select this project BAM EXPLORATION SECTION Bam Exploration Interface COUNT SECTION PRE ANALYSIS SECTION RNASeqGUI Read Count Interface is ready to work on Myproject project Select a bam folder Open Provide a GTF Open Counting Mode Union Ignore Strand Save outputs Save Results Name CSV Start Counting Reads How touse this Interface Count Reads Figure 11 Read Count Interface Gene Id control_l control_2 treated_l1 treated_2 ENSG00000000003 455 463 583 598 ENSG00000000005 0 0 0 1 ENSG00000000419 1174 1210 1545 1533 ENSG00000000457 260 256 305 349 ENSG00000000460 550 607 709 741 Figure 12 An example of a count file with 20062 genes The row names are given by the Gene Id in the annotation file gtf the column names are given by the alignment file names the bam files 22 BAM CAPLURKALION gt E HON Bam Exploration Interface COUNT SECTION Read Count Interface PRE ANALYSIS SECT
45. rmat Workspace OAEI DF Misc Window Help Package Manager Package Installer 1 The Package Installer Requires An Internet Connection Data Manager In the Package Installer pull down the top left menu and select Local Binary Package eoo R Package Installer Packages Repository CRAN binaries CRAN sources BioConductor binaries Packages CE f ckage rct y BioConductor sources Other Repository Other Directory URL Local Binary Package Local Source Package Local Package Directory installed Version Repository Version Install Location At System Level in R framework nstall Selected At User Level Install Dependencies In Other Location Will Re Asked Linan Installation Figure 3 Under Package and Data pull down menu select Package In staller and navigate to where you have downloaded the binary package From 10 install packages digest install packages ggplot2 install packages RColorBrewer install packages VennDiagram install packages XML 3 Type the order of the list below is important source http bioconductor org biocLite R biocLite biomaRt biocLite DEXSeq biocLite pasilla biocLite GenomicRanges biocLite GenomicFeatures biocLite Rsamtools biocLite edgeR biocLite baySeq biocLite NOISeq biocLite DESeq biocLite DESeq2 biocLite gplots biocLite EDASeq biocLite leeBamViews biocLite preprocess
46. ry layout as reported in Figure BI We choose a 0 05 value as the Padj Finally we click on Run DESeq button The DESeq analysis is performed and two result text files are created and saved in the Results folder We click on NOISeq button In the NOISeq Analysis Interface we select the 2L_counts csv count file We type the T T T U U U U sequence in the Factors field We type T1 T3 T4 U1 U3 U4 U6 in the TissueRun field to specify the library layout as specified in Figure BI We select biological in the Replicate field We choose a 0 6 value as the prob Finally we click on Run NOISeq button The NOISeq analysis is performed and two result text files are created and saved in the Results folder Once all the results have been obtained we can start inspecting them by clicking on Result Inspection Interface We click on EdgeR DESEq and NOISeq buttons at the same time At each click we can see the Result Inspection Interface growing see the top right of the Figure 44 For each method we select the corresponding result file by giving the all path to the file in the Select File field and we click on Plot FC on FDR Hist and on Volcano Plot of each method We also provide a gene id to display a specific gene in this case we type FBgn0000559 in the Gene Id field as shown in Figure 44 that is the most expressed gene found in the heatmap in Figure 3 Finally we compare the results by clicking on Result Comparison Interface We fill all th
47. s This func tion performs the Upper Quartile Bullard et al 2010 normalization e TMM makes use of tmm function of the NOISeq package Tarazona et al 2011 This function takes a count file as specified in Figure 12 and returns a count file with normalized numbers This function performs the TMM Robinson et al 2010 normalization 26 neau suun unerarte PRE ANALYSIS SECTION Data Exploration Interface Normalization Interface DATA ANALYSIS SECTION Data amay Interface RNASeqGUI Result Inspection please choose one of the methods below to identify DE genes You are working on MyProject project EdgeR DESeq DESeq2 Contacts and Il NoiSeq BaySeq Figure 15 Data Analysis Interface e Full Quantile makes use of normalize quantiles function of the preprocessCore package This function takes a count file as specified in Figure 12 and returns a count file with normalized numbers This function performs the Full Quantile Bolstad et al 2003 Smyth et al 2005 normalization 5 5 Data Analysis Section This section contains the Data Analysis Interface shown in Figure I5 and represents the core of RNASeqGUI This interface includes five different sta tistical methods to detect differentially gene expression such as EdgeR DESeq DESeq2 NoiSeq BaySeq 5 5 1 EdgeR e The EdgeR method Robinson et al 2007 Robinson et al 2008 Robinson et al 2010 McCarthy et al 2012 see Figure 6 ta
48. s about the identification of the differentially expressed genes that can be performed by several methods such as DESeq DESeq2 EdgeR NOISeq BaySeq Finally the fifth section regards the inspection of the results produced by these methods and the quantitative comparison among them This software is not just a collection of some known methods and functions but it is designed to guide the user during the entire analysis process Moreover the GUI is also helpful for those who are expert R users since it speeds up the usage of the included RNASeq methods drastically Current implementation allows to handle the simple experimental design where the interest is on the experimental condition future work will cover complex designs PE E E Gearon 1131 Napoli Italy f web development by bryant smith Figure 1 The http bioinfo na iac cnr it RNASeqGUI web page 2 2 For MacOS users 1 Install Xcode developer tools at least version 5 0 1 from Apple Store it is free 2 Install XQuartz 2 7 5 dmg from http xquartz macosforge org landing 3 Install GTK_2 24 17X11 pkg from http r research att com 4 Then go to Section 3 2 3 For Windows users 1 download gtk bundle_2 22 1 20101229 win64 zip from http ftp gnome org pub gnome binaries win64 gtk 2 22 2 This is a bundle containing the GTK stack and its dependencies for Windows To use it create some empty folder like C opt gtk 3 Unzip this bundle 4
49. s available in two modes By default the log of the counts we add 1 to each number in the count file to avoid the problem of log 0 will be used to generate the density function It is possible to uncheck this mode by clicking in the log check box see Figure I3 The MDPlot makes use of MDplot function of the EDASeq package Risso et al 2011 This function takes a count file and two integers Columni and Column2 and returns a plot showing the mean of the two selected columns against their difference gene by gene For this function the log check box does not change anything The MeanVarPlot makes use of meanVarPlot function of the EDASeq package Risso et al 2011 This function takes a count file and returns a plot showing the mean of all columns found in the file against the 24 variance gene by gene For this function the log check box does not change anything The Heatmap makes use of heatmap function of the stats package This function takes a count file and an integer N in the How many genes in the Heatmap field The function returns an heat map of the N most expressed genes on average The columns of the heatmap are the samples while the rows in the heat map represent the gene ids of the most expressed ones An example of heat map is shown in Figure Moreover for this function it is possible to generate the heatmap either of the raw counts or the log of the counts we add 1 to each number in the count file to avo
50. saved in the Plots folder 5 5 4 NoiSeq The NoiSeq Tarazona et al 2011 method see Figure takes an input count file as the one shown in Figure 2 via the Open button and returns two text files 32 padj 0 954 0 911 0 768 0 937 0 994 Please choose one of the methods below to identify DE genes You are working on MyProject project EdgeR DESeq DESeq2 569 BaySeq RNASeqGUI NoiSeq Interface is ready to work on MyProject project Select a count file Open Headers M whatis the Column Separator Tissue TissueRun Replicate technical Prob 0 8 Save outputs Save Results Name cSV How to use this interface Run NoiSeq Figure 23 NoiSeq Interface The first text file shows the results of this method see Figure BA where M is the log2 ratio of the two conditions The second text file shows the differentially expressed genes only 2 The first file is named with the name of the input file plus _results_Noiseq txt suffix The output count file is saved with the name specified by the user in the Name field see Figure B3 If no name is specified by the user then the second file is named with the name of the input file plus _prob 0 8_DE_genes_Noiseq txt suffix where 0 8 is the chosen pos terior probability for rejection Both text files are saved in the Results folder Both plots are saved in the Plots folder 5 5 5 BaySeq e The BaySeq Hardcastle et al 2010 metho
51. t the 2L_counts csv count file We type the T T T U U U U sequence in the Factors field to specify which are the treated samples called T and which are the untreated ones called U as reported in Figure We choose a 0 05 value as the FDR Finally we click on Run EdgeR button The EdgeR analysis is performed and two result text files are created and saved in the Results folder 46 12 10 Count Distribution 1 L T T T T T T T Data Bam e1 2L_1 bam Data Bam e1 2L_3 bam Data Bam e1 2L_4 bam Data Bam e1 2L_U1 bam Data Bam e1 2L_U3 bam Data Bam e1 2L_U4 bam Figure 39 Box plot generated by the BoxPlot function 47 Data Bam e1 2L_U6 bam Full Quantile BoxPlot l iii Oo T T T T T T T 1 2 3 4 5 6 7 12 6 4 Figure 40 Boxplot of the counts shown in Figure 39 after the full quantile normalization Count Plots 0 60000 0 100000 0 100000 petit tt iL 4 l 1000 000 4 o q q o S jata Bam e1 2L1 4 o o o o o o s insi 2060 og oo 06 Q 60 Bo id d o 7 o g q a d 8 o o 9 o 9 o 2 o 9 So Oo lata Bam e1 2L 3 4 o o o o o o o PA s Pg o kej q v v d o o o q o o o o o o z P o fata Bam et 2L_4 4 06 oo So JE 1 E I o o og lo a S 8 oo 00 cc 5 v o v of o9 Y o g 8 ta Bam e1 2L_U3 6 o ri oF o o o 8 pla Bam e1 2L_U4
52. the interface will grow and we get the interface shown in Figure 28 Therefore for each method we have Plot FC FDR Hist or P value Hist and Volcano Plot functions except for the BaySeq method since this method already provides an MAplot and a PosteriorPlot during the analysis process that can be run in the BaySeq Analysis Interface 36 RNASeqGUI Result Inspection interface is ready to work on MyProject project Select a result file onon Headers amp Column Separator Gene Id FDR 0 05 Padj 0 05 Prob 0 9 EdgeR BaySeq 26569 DESeq2 NoiSeq EdgeR Analysis PlotFC FDRHist Volcano Plot BaySeq Analysis FDR Hist Likelihood Hist DESeq Analysis PlotFC P value Hist Volcano Plot DESeq2 Analysis PlotFC P value Hist Volcano Plot NoiSeq Analysis Plot FC ProbHist Volcano Plot Figure 28 Result Inspection Interface after clicking all the five buttons at the top 37 DATA ANALYSIS SECTION Data Analysis Interface POST ANALYSIS SECTION Result Inspection Interface Result Comparison Interface RNASeqGUI Result Comparison Interface is ready to work on Myproject project Select the First file to Compare Open Headers M Column Separator First label Select the Second file to Compare Open Headers Column Separator Second label Selectthe Third file to Compare Open Headers amp Column Separator Third label How to
53. therwise we get the following error Error in as vector x character cannot coerce type environment to vector of type character If the user wants to use these functions in this case the 2L_1 bam file must be temporary removed from the demo folder before using them In this case if we use those functions without the 2L_1 bam file we get the plots in Figure 32 and in Figure B3 respectively Subsequently we click on Read Count Interface and select the bam folder demo and the 2L_Drosophila_melanogaster BDGP5 70 gtf annotation file We select Union as Counting Mode and check the Ignore Strand box as shown in Figure 6 Hence we click on Count Reads button As result of 42 Data Bam demo 2L_3 bam C____ 1060000000000000000 ___ 10000000000000000000000000 1006000000000000000000000000000 C_ 1000000000000000000000000000000000 10 000000000000000000000000000000000000000 0000000000000000000000000000000000000000 1 0o000000000000000000000000000000000000000 1 0000000000000000000000000000000000000000 1 0000000000000000000000000000000000000000 10000000000000 000000000000000000000000000 1 0000000000000000000000000000000000000000 0000000000000 000000000000000000000000000 1 0000000000000 000000000000000000000000000 1 o000000000000000000000000000000000000000 10000000000000 000000000000000000000000000 10000000000000 000000000000000000000000000 0
54. user project This report reports the session information that describes all used package versions by RNASeqGUI at the time of the project creation along side with the name of the project time date and the parameters fdr padj etc the user selected during the usage of the GUI 39 RNASeqGuUI Home Example Manual Download Contact Material Credits Figure 30 At http bioinfo na iac cnr it RNASeqGUI Example we can download the example 6 Usage Example We can start using RNASeqGUI by downloading the example data at http bioinfo na iac cnr it RNASeqGUI Example as shown in Figure We download the folder called example_RNASeqGUI tar gz we extract this bundle and open it Inside this we find a folder called demo a gtf file called 2L_Drosophila_ melanogaster BDGP5 70 gtf and a text file called README txt file 6 1 Data Preparation In this usage example we start the analysis of the RNA Seq data from bam files and we compare the results of EdgeR DESeq and NOISeq against each other We downloaded the dataset published by Brooks et al 2011 This dataset has already been used in Anders et al 2013 as a real data working example We downloaded the data from http www ncbi nlm nih gov sra term SRP001537 by following the instructions described in Anders et al 2013 at the page 1771 The entire experiment is available at http www ncbi nlm nih gov geo query acc cgi acc GSE18508 The dataset consists o
55. zarry R A Astrand M SpeedT P 2003 A Comparison of Normalization Methods for High Density Oligonucleotide Array Data Based on Bias and Variance Bioinfor matics 19 2 185 193 Brooks et al 2011 Brooks A N Yang L Duff M O Hansen K D Park J W Dudoit S Brenner S E Graveley B R 2011 Conserva tion of an RNA regulatory map between Drosophila and mammals Genome Research 21 193 202 Bullard et al 2010 Bullard J H Purdom E Hansen K D Dudoit S 2010 Evaluation of statistical methods for normalization and dif ferential expression in mRNA seq experiments BMC Bioinformatics 11 94 Hardcastle et al 2010 Hardcastle T J Kelly K A 2010 baySeq Empir ical Bayesian methods for identifying differential expression in se quence count data Bioinformatics 11 422 Kim et al 2013 Kim D Pertea G Trapnell C Pimentel H Kelley R SalzbergS L 2013 TopHat2 accurate alignment of transcriptomes in the presence of insertions deletions and gene fusions Genome Biology 14 R36 Lawrence et al 2010 Lawrence M Temple Lang D 2010 RGtk2 A Graphical User Interface Toolkit for R Journal of Statistical Soft ware 37 8 Lawrence et al 2013 Lawrence M Huber W Pags H Aboyoun P Carl son M 2013 Software for Computing and Annotating Genomic Ranges PLoS Comput Biol 9 8 58 Lohse et al 2012 Lohse M Bolger A M Nagel A Fernie A R Lunn J E
Download Pdf Manuals
Related Search