Courses in R and a bit more from Aedin Culhane

Intro to Bioconductor

Introduction to R and Bioconductor, May 2012

Course Description

This course is an introduction to R and Bioconductor, a powerful and flexible statistical language for analysis of genetic and genomics data (http://www.bioconductor.org/). The course will introduce attendees to the basics of using R for statistical programming, computation, graphics, and modeling, especially for analyzing high-throughput genomic data. We will start with a basic introduction to the R language, reading and writing data, and plotting data. Case studies and data will all be based on real gene expression and genomics data. We will introduce the main classes and packages in Bioconductor. Our goal is to get attendees up and running with R and Bioconductor such that they can use it in their research and are in a good position to expand their knowledge of R and Bioconductor on their own.

Instructors

Aedin Culhane contact: aedin@jimmy.harvard.edu
Guest Lecturers:
- Nicolai Juul Birkbak (SNP data)
- Svitlana Tyekucheva, Alejandro Quiroz (Gene Set Analysis)
- J. Fah Sathirapongsasuti (Exome Seq CNV)

Schedule

May 16th- 9:15am – 5:00pm
May 17th – 9:15am – 12:00pm

Required Software

I recommend the following software.

R. Download R from from the R home page
The integrated development envirnoment (IDE) R Studio available for Windows, Mac or Linux OS
Install Bioconductor. Start R and type the following command (it requires an internet connection)
```
source("http://www.bioconductor.org/biocLite.R")
biocLite()
```
Latex
- Windows: MikTex or there is a easy-to-install Tex software bundle called proTeXt which includes MikTex, the latex editor TeXnicCenter and Ghostscript
- MacOS: MaxTEx
Latex Editor ( a comparison of editors )
- Windows: Either TexWorks, TeXnicCenter or simply an enchanced notepad like Notepad++
- MacOS: TeXshop
- Linux: I tend to use either Kate (within KDE), Emacs or Texworks which is cross platform
- Convert Tex to a MSword Document using TeX4ht

These should be pretty straightforward download and install, but this document provides a little more detail instruction on installing R and Bioconductor (from May 2011) extension packages (not required for course).

Agenda

Please install the following packages by copying and pasting this into R, or sourcing this script.

install.packages(c("R2HTML", "lme4", "manipulate", "RColorBrewer", "googleVis","network","igraph","lattice","gplots", "venneuler")

Day 1

History and Background to R, Installing R website slides
Introduction to R language (classes, subsetting etc)
Plotting in R

Reading Data into R and Bioconductor

## Install Bioconductor Packages
source("http://www.bioconductor.org/biocLite.R")
biocLite(c("GEOquery", "ArrayExpress", "frma", "arrayQualityMetrics, "affycoretools", "made4"))
biocLite("HGU95Av2_Hs_ENSG",respos="http://brainarray.mbni.med.umich.edu/bioc")

Expression Set Class, S4 Class,Example data workflows in R

slides

Function to create an ExpressionSet given 2 matrices (or data.frames) containing 1) expression data and 2) annotation

makeEset<-function(eSet, annt){
          #Creating an ExpressionSet from eSet, a normalized gene expression matrix
          # and annt, a data.frame containing annotation
    metadata <- data.frame(labelDescription = colnames(annt), row.names=colnames(annt))
    phenoData<-new("AnnotatedDataFrame", data=annt, varMetadata=metadata)
    if (inherits(eSet, "data.frame")) eSet= as.matrix(eSet)
    if (inherits(eSet, "ExpressionSet")) eSet=exprs(eSet)
    data.eSet<-new("ExpressionSet", exprs=eSet, phenoData=phenoData)
    print(varLabels(data.eSet))
    return(data.eSet)
}

Normalization and custom cdf files Install packages for this tutorial
Clustering and Exploratory Data Analysis in R slides
- Notes on EDA in R
- R code for EDA Notes
- data.vsn
- annt.txt
- Other potentially useful packages EDA for sequence data EDASeq
Feature Selection (Limma etc), GSEA and Annotating data using annotate/BiomaRt
- Feature Selection and Gene Annotation Notes on Feature Selection using Limma and Annotating Genes in R
- R code for Feature Selection/Annotation Notes
- References which compare different feature selection approaches
  - Jeffery IB, Higgins DG, Culhane AC. (2006) Comparison and evaluation of microarray feature selection methods. BMC Bioinformatics 7:359.
  - Murie C, Woody O, Lee AY, Nadon R. (2009) Comparison of small n statistical tests of differential expression applied to microarrays. BMC Bioinformatics. 10:45.
- Gene Annotation HTML output Results of aafTableAnn()
Gene Set Analysis For this analysis you will need the following packages
```
source("http://bioconductor.org/biocLite.R")
biocLite("RTopper")
biocLite("limma")
```
- GeneSigDB http://compbio.dfci.harvard.edu/genesigdb
- MSigDB http://www.broadinstitute.org/gsea/msigdb/index.jsp
- KEGG http://www.genome.jp/kegg/
- GO http://www.geneontology.org/
Literature mining, tag cloud (will not be presented during class)
- See the Text Mining Package R library tm and wordcloud . There are also many network and graph packages will be covered later
- Sample R script to create a word cloud from frequently occuring words in pubmed abstracts
- Example of script output
- Also see packages annotate, XML etc
1. Survival Analysis (Survcomp, Ben)
  A few Code Tips
  
  Function to create an ExpressionSet given 2 data matrices (or data.frames) containing 1) expression data and 2) annotation
```
makeEset<-function(eSet, annt){
          #Creating an ExpressionSet from eSet, a normalized gene expression matrix
          # and annt, a data.frame containing annotation
    metadata <- data.frame(labelDescription = colnames(annt), row.names=colnames(annt))
    phenoData<-new("AnnotatedDataFrame", data=annt, varMetadata=metadata)
    if (inherits(eSet, "data.frame")) eSet= as.matrix(eSet)
    if (inherits(eSet, "ExpressionSet")) eSet=exprs(eSet)
    data.eSet<-new("ExpressionSet", exprs=eSet, phenoData=phenoData)
    print(varLabels(data.eSet))
    return(data.eSet)
}
```
  Starting with Biomart
```
 
library(biomaRt)
mart=useMart("ensembl")
mart<-useDataset("hsapiens_gene_ensembl",mart)
geneAnnt<-getBM(attributes=c("affy_hg_u95av2","hgnc_symbol","chromosome_name","band", "entrezgene"),filters="affy_hg_u95av2",values=c("1939_at","1503_at","1454_at"), mart=mart)
```
  Updated.May 2012. Aedin Culhane