%\VignetteEngine{knitr::knitr} %% \VignetteIndexEntry{01 Project Overview -- Slides} \documentclass[xcolor=dvipsnames]{beamer} \usepackage{BioconductorSlides} \hypersetup{colorlinks,linkcolor=,urlcolor=Blue} \AtBeginSection[] { \begin{frame}{Outline} \tableofcontents[currentsection] \end{frame} } \begin{document} \title{R / Bioconductor for Analysis and Comprehension of High-Throughput Sequence Data} \author{Martin T. Morgan (\url{mtmorgan@fhcrc.org}) \\ Fred Hutchinson Cancer Research Center \\ Seattle, WA, USA} \date{3 February 2014} \maketitle \section*{Introduction} \begin{frame}{Overview} \begin{enumerate} \item Introduction to \R{} and \Bioconductor{} \item Sequencing work flows \item Successful computational biology software \item Exemplars: algorithms into actions \item Challenges \& opportunities \end{enumerate} \end{frame} \begin{frame}{Introduction: \Bioconductor} Analysis and comprehension of high-throughput genomic data \begin{itemize} \item \url{http://bioconductor.org} \item $>11$ years old, $749$ packages \end{itemize} Themes \begin{itemize} \item Rigorous statistics \item Reproducible work flows \item Integrative analysis \end{itemize} \end{frame} \begin{frame}{Introduction: \Bioconductor} \begin{columns} \column{.5\textwidth} \begin{itemize} \item 1341 PubMed full-text citations in trailing 12 months \item 28,000 web visits / month; 75,000 unique IP downloads / year \item Annual conferences; courses; active mailing list; \ldots \end{itemize} \column{.5\textwidth} \includegraphics[width=\textwidth]{figures/usage-geography} \end{columns} \bigskip\par \textbf{\Bioconductor{} Conference}, July 30 - Aug 1, Boston, USA \end{frame} \begin{frame}{Introduction: What is \Bioconductor{} good for?} \begin{itemize} \item Microarrays: expression, copy number, SNPs, methylation, \ldots \item Sequencing: RNA-seq, ChIP-seq, called variants, \ldots \begin{itemize} \item Especially \emph{after} assembly / alignment \end{itemize} \item Annotation: genes, pathways, gene models (exons, transcripts, etc.), \ldots \item Flow cytometry, proteomics, image analysis, high-throughput screens, \ldots \end{itemize} \end{frame} \section*{Sequencing work flows} \begin{frame}[fragile]{Introduction: \R} \begin{columns} \column{.45\textwidth} \begin{itemize} \item \url{http://r-project.org} \item Open-source, statistical programming language; widely used in academia, finance, pharma, \ldots \item Core language and base packages \item Interactive sessions, scripts \item $>5000$ contributed packages \end{itemize} \column{.55\textwidth} <>= ## Two 'vectors' x <- rnorm(1000) y <- x + rnorm(1000, sd=.5) ## Integrated container df <- data.frame(X=x, Y=y) ## Visualize plot(Y ~ X, df) ## Regression; 'object' fit <- lm(Y ~ X, df) ## Methods on the object abline(fit) # regression line anova(fit) # ANOVA table @ \end{columns} \end{frame} \begin{frame}[fragile]{Sequencing: Work flows} \begin{columns} \column{.5\textwidth} \begin{enumerate} \item Experimental design \item `Wet lab' sample prep \item Sequencing \begin{itemize} \item 100's of millions of reads \item 30-150 nucleotides \item Single and paired-end \item Bar codes, lanes \& flow cells \end{itemize} \item Alignment \item Analysis: DNA, RNA, epigenetics, integrative, microbiome, \ldots \end{enumerate} \column{.5\textwidth} \includegraphics[width=\textwidth,height=!]{figures/Solexa-bridge-pcr.jpg} \par Bentley et al., 2008, Nature 456: \href{http://www.ncbi.nlm.nih.gov/pubmed/18987734}{53-9} \end{columns} \end{frame} {\scriptsize\begin{verbatim} @ERR127302.1703 HWI-EAS350_0441:1:1:1460:19184#0/1 CCTGAGTGAAGCTGATCTTGATCTACGAAGAGAGATAGATCTTGATCGTCGAGGAGATGCTGACCTTGACCT + HHGHHGHHHHHHHHDGG>CE?=896=: @ERR127302.1704 HWI-EAS350_0441:1:1:1460:16861#0/1 GCGGTATGCTGGAAGGTGCTCGAATGGAGAGCGCCAGCGCCCCGGCGCTGAGCCGCAGCCTCAGGTCCGCCC + DE?DD>ED4>EEE>DE8EEEDE8B?EB<@3;BA79?,881B?@73;1?######################## @ERR127302.1705 HWI-EAS350_0441:1:1:1460:13054#0/1 AAAACACCCTGCAATCTTTCAGACAGGATGTTGACAATGCGTCTCTGGCACGTCTTGACCTTGAACGCAAAG + EEDEE>AD>BBGGB8E8EEEGBGGGGBGGGGG3G>E3*?BE??BBC8GB8??:??GGDGDDD>D>BGGD8EG,<6D@<@G@>AB@B?8AA>CE@D8@B=?CC>AG @ERR127302.1707 HWI-EAS350_0441:1:1:1461:6983#0/1 CGACGCTGACACCGGAACGGCAGCAGCAGCAGGACGATTAAGACAAGGAGGATGGCTCCACAGACGCTCATG + GEEGEGE@GGGGGGEGGGGGBB>G3?33?8*;;79?<9@?DD8@DDEE888;-BB?.A############## @ERR127302.1708 HWI-EAS350_0441:1:1:1461:10827#0/1 AAAGAAGGTCCTTGCAATAGACTGCCTCTGCTTGAGAACTTATGATGTAATTATTGCATGCTGCTAATATAC + GGGGGDDEBFGGGGGBE,DAGDDGGGEEEGACEBEFDEEFEDH:@.7@49;88G8>G>DDG@D>D@G@GE>@DDBDDG>= ## Use the 'ShortRead' package library(ShortRead) ## Create an object to represent a sample from a file sampler <- FastqSampler("ERR127302_1.fastq.gz") ## Apply a method to yield a random sample fq <- yield(sampler) ## Access sequences of sampled reads using `sread()` ## Summarize nucleotide use by cycle ## 'abc' is a nucleotide x cycle matrix of counts abc <- alphabetByCycle(sread(fq)) ## Subset of interesting nucleotides abc <- abc[c("A", "C", "G", "T", "N"),] @ \end{frame} \begin{frame}[fragile]{Sequencing: The \Biocpkg{ShortRead} package} \begin{columns} \column{.5\textwidth} <>= ## Create a plot from a ## matrix matplot(t(abc), type="l", lty=1, lwd=3, xlab="Cycle", ylab="Count", cex.lab=2) ## Add a legend legend("topright", legend=rownames(abc), lty=1, lwd=3, col=1:5, cex=1.8) @ \column{.5\textwidth} \includegraphics[width=\textwidth]{figures/abc} \end{columns} \end{frame} \begin{frame}{Sequencing: Essential packages and classes} \begin{itemize} \item \Biocpkg{Biostrings} and \Rclass{DNAStringSet} \item \Biocpkg{GenomicRanges} and \Rclass{GRanges} \item \Biocpkg{GenomicFeatures} and \Rclass{TranscriptDb} \item \Biocpkg{VariantAnnotation} and \Rclass{VCF} \item Input and output: \Biocpkg{rtracklayer} (WIG, BED, etc.), \Biocpkg{Rsamtools} (BAM), \Biocpkg{ShortRead} (FASTQ) file input \end{itemize} \end{frame} \section*{Principles} \begin{frame}{Principles: Some key points} \begin{itemize} \item \R{} is a high-level programming language, so lots can be accomplished with just a little code \item Packages such as \Biocpkg{ShortRead} provide a great way to benefit from the expertise of others (and to contribute your own expertise back to the community!) \begin{itemize} \item The path from `user' to `developer' is not that long, and has been taken by many! \end{itemize} \item Objects and methods such as \Rclass{data.frame}, \Rclass{ShortReadQ} and \Rcode{alphabetByCycle()}) help to manage complicated data \begin{itemize} \item Reducing possibility for clerical and other mistakes \item Facilitating inter-operability between different parts of an analysis \end{itemize} \item Scripts make work flows reproducible \item Visualizing data is an important part of exploratory analysis \end{itemize} \end{frame} \begin{frame}{Principles: Successful computational biology software} \begin{enumerate} \item Extensive: software, annotation, integration \begin{itemize} \item 750 inter-operable \Bioconductor{} packages \end{itemize} \item Statistical: volume, technology, experimental design \begin{itemize} \item \R{} a `natural' for statistical analysis \end{itemize} \item Reproducible: long-term, multi-participant science \begin{itemize} \item Objects, scripts, vignettes, packages, \ldots encourage reproducible research \end{itemize} \item Leading edge: novel, technology-driven \begin{itemize} \item Packages and user community closely track leading edge science \end{itemize} \item Accessible: affordable, transparent, usable \begin{itemize} \item \Bioconductor{} is free and open, with extensive documentation and an active and supportive user community \end{itemize} \end{enumerate} Case study: differential expression of known genes; see also \href{https://bioconductor.org/help/course-materials/2013/EMBOBGI/reproducible-research.pdf}{reproducible research} lecture. \end{frame} \section*{Exemplars} \begin{frame}{Exemplars: Algorithms to action} \begin{enumerate} \item Batch effects \item Methylation \item RNA-seq Differential Representation \item Visualization \end{enumerate} \end{frame} \begin{frame}{Exemplar: Differential Representation} Haglund et al., 2012 \href{http://www.ncbi.nlm.nih.gov/pubmed/23024189}{J Clin Endocrin Metab} \bigskip\par \begin{columns} \column{.5\textwidth} \begin{itemize} \item Scientific finding: identify genes whose expression is regulated by estrogen receptors in parathyroid adenoma cells \item Statistical challenges: between-sample normalization; appropriate statistical model; efficient estimation; \ldots \end{itemize} \column{.5\textwidth} \includegraphics[width=\textwidth]{figures/DESeq2_parathyroid-plotMApadjchange.png} \end{columns} \bigskip\par\Bioconductor{} support: \Biocpkg{DESeq2}, \Biocpkg{edgeR}, many statistical `lessons learned' from microarrays; extensive integration with down-stream tools \end{frame} \begin{frame}{Exemplar: Batch Effects} Leek et al., 2010, Nature Reviews Genetics 11, \href{http://www.nature.com/nrg/journal/v11/n10/abs/nrg2825.html}{733-739}, Leek \& Story \href{http://dx.doi.org/10.1371/journal.pgen.0030161}{PLoS Genet 3(9): e161} \begin{columns} \column{.5\textwidth} \begin{itemize} \item Scientific finding: pervasive batch effects \item Statistical insights: surrogate variable analysis: identify and build surrogate variables; remove known batch effects \item Benefits: reduce dependence, stabilize error rate estimates, and improve reproducibility \end{itemize} \Bioconductor{} support: \Biocpkg{sva} \column{.5\textwidth} \only<1>{ \includegraphics[width=\textwidth]{figures/nrg2825-f2.jpg} \par{\small HapMap samples from one facility, ordered by date of processing. From } } \only<2>{ \begin{enumerate} \item Remove signal due to variable(s) of interest \item Identify subset of genes driving orthogonal signatures of EH \item Build a surrogate variable based on full EH signature of that subset \item Include significant surrogate variables as covariates \end{enumerate} EH: expression heterogeneity } \end{columns} \end{frame} \begin{frame}{Exemplar: Methylation} Hansen et al., 2011, Nature Genetics 43, \href{http://www.nature.com/ng/journal/v43/n8/full/ng.865.html}{768-775} \begin{itemize} \item Scientific finding: stochastic methylation variation of cancer-specific de-methylated regions (DMR), distinguishing cancer from normal tissue, in several cancers. \item Statistical challenges: smoothing, non-specific filtering, $t$ statistics, find DMRs \end{itemize} \bigskip\par \includegraphics[width=\textwidth]{figures/bsseq_analysis-1.png} \medskip\par \Bioconductor{} support: whole-genome (\Biocpkg{bsseq}) or reduced representation (\Biocpkg{MethylSeekR}) bisulfite sequencing; Illumina 450k arrays (\Biocpkg{minfi}) \end{frame} \begin{frame}{Exemplar: Visualization} \begin{columns} \column{.5\textwidth} \Biocpkg{Gviz}\par \only<1-2>{ \begin{itemize} \item Track-like visualizations \item Data panels \item Fully integrated with \Bioconductor{} sequence representations \end{itemize} } \Biocpkg{ggbio}\par \only<3>{ \begin{itemize} \item Comprehensive visualizations \item \Rfunction{autoplot} file and data types \item Fully integrated with \Bioconductor{} sequence representations \end{itemize} } \Biocpkg{epivizr}\par \only<4>{ \begin{itemize} \item Genome browser with socket communication to \R{} \item Fully integrated with \Bioconductor{} sequence representations \end{itemize} } \column{.5\textwidth} \only<1>{\includegraphics[width=\textwidth]{figures/Gviz-vignette-1.png}} \only<2>{\includegraphics[width=\textwidth]{figures/Gviz-vignette-2.png}} \only<3>{\includegraphics[width=\textwidth]{figures/ggbio-vignette-1.png}} \only<4>{\includegraphics[width=\textwidth]{figures/epivisr.png}} \end{columns} \end{frame} \section*{Challenges \& Opportunities} \begin{frame}{Challenges \& Opportunities} \begin{itemize} \item Big data -- transparent management within \R, facile use of established resources \item Developer and user training \end{itemize} Resources \begin{itemize} \item \url{http://r-project.org}, \emph{An Introduction to \R} manual; Dalgaard, \emph{Introductory Statistics with \R}; \href{http://rfordummies.com/}{\R{} for Dummies} \item \url{https://bioconductor.org/} \item \url{http://rstudio.org} \item \href{http://stackoverflow.com/questions/tagged/r}{StackOverflow}, \Bioconductor{} \href{https://bioconductor.org/help/mailing-list/mailform/}{mailing list} \end{itemize} \end{frame} \section*{Acknowledgements} \begin{frame}{Acknowledgements} \begin{itemize} \item \Bioconductor{} team: Marc Carlson, Valerie Obenchain, Herv\'e Pag\`es, Paul Shannon, Dan Tenenbaum \item Technical advisory council: Vincent Carey, Wolfgang Huber, Robert Gentleman, Rafael Irizzary, Sean Davis, Kasper Hansen \item Scientific advisory board: Simon Tavar\'e, Vivian Bonazzi, Vincent Carey, Wolfgang Huber, Robert Gentleman, Rafael Irizzary, Paul Flicek, Simon Urbanek. \item NIH / NHGRI U41HG0004059 \item The \Bioconductor{} community \item \ldots and the organizers of this course! \end{itemize} \end{frame} \end{document}