This course provides a broad introduction to the analysis and comprehension of high-throughput sequence data using R and Bioconductor; many other useful software tools are not discussed. The topics and learning objectives to be address are as follows:

• Introduction to Bioconductor: Familiarity with overall sequence analysis work flows. The diversity of analysis challenges addressed by Bioconductor. The role of Bioconductor packages in specific work flow steps. Bioconductor resources for help and effective use.

• Next Generation Sequencing data manipulation: Common genomic file types (FASTA/FASTQ, BED, WIG, GTF, etc) and their Bioconductor representation and manipulation. Essential Bioconductor data structures (e.g., DNAStringSet, GenomicRanges). Bioconductor gene and genome 'annotation' resources. Simple Bioconductor solutions to common bioinformatic tasks. Primary packages: GenomicFiles, GenomicRanges, GenomicFeatures, ShortRead, Biostrings.

• Variant calling and annotation: Conceptual understanding of variant calling approaches and (primarily non-Bioconductor) tools. Exploratory approaches to variant calling in Bioconductor. Working with called variants, e.g., classifying variants to genomic regions of interest and predicting effects of variants. Primary packages: VariantTools, r5vc, VariantAnnotation.

• RNA-seq differential expression analysis: Bioconductor known-gene RNA-seq differential expression work flow, from aligned reads to differential expression of genes. Important statistical issues and their resolution. Placing results of differential expression analysis into biological context. Brief discussion of novel-gene and transcript-level RNAseq differential expression analysis. Primary packages: DESEq2, edgeR.

• Working with large data: Overcoming common pitfalls in memory management and computational throughput when writing R code. Adopting restriction, iteration, and parallel evaluation strategies for working with large data. Primary packages: BiocParallel, GenomicFiles.

• Data visualization: Challenges to visualizing genomic data. Opportunities for familiar and novel genomic visualizations in Bioconductor. Approaches to interactive visualization with shiny. Adopting strategies for reproducible and collaborative research. Primary packages: Gviz, ReportingTools, (ggbio).

Additional topics may be discussed briefly and on an ad hoc basis.

Tentative Schedule

Monday

9:00 - 12:30 (break 10:30 - 11:00)

• Introduction to R and Bioconductor
  • From vectors to objects
  • Help! and other resources
  • Why Bioconductor? SummarizedExperiment

1:30 - 5:00 (break 3:00 - 3:30)

• Working with sequence data in Bioconductor

  • Overall sequence work flows
  • Key Bioconductor packages and classes -- GenomicRanges, Biostrings, ShortRead, rtracklayer
  • 'Annotation' resources -- GenomicFeatures, AnnotationHub, biomaRt
  • Working with large data -- BiocParallel, GenomicFiles
  • Common bioinformatics tasks, easy Bioconductor solutions

• Participant lightning talks -- short, ad hoc or lightly prepared presentations by workshop participants to introduce their work. A couple of participants will volunteer during each time slot throughout the course.

Tuesday

9:00 - 12:30 (break 10:30 - 11:00)

• Variant annotation and calling

  • Working with called variants: VCF files, VariantAnnotation, VariantFiltering.
  • Calling variants. Common workflows outside Bioconductor. Alternative approaches in Bioconductor. VariantTools, r5vc.

• Participant lightning talks -- short, ad hoc or lightly prepared presentations by workshop participants to introduce their work.

1:30 - 5:00 (break 3:00 - 3:30)

• RNA-seq differential expression.

  • Overall work flow, statistical issues, and implementation using DESeq2 and edgeR
  • kallisto and other newer approaches; limma voom()

• Participant lightning talks -- short, ad hoc or lightly prepared presentations by workshop participants to introduce their work.

Wednesday

9:00 - 12:30 (break 10:30 - 11:00)

• One of:

  • Working with large data GenomicFiles, BiocParallel
  • csaw, ChIPseeker and other ChIP-seq analysis
  • Gene set and other down-stream analysis facilities. limma gene set functions, Category, GOstats, WGCNA.

• Participant lightning talks -- short, ad hoc or lightly prepared presentations by workshop participants to introduce their work.

1:30 - 5:00 (break 3:00 - 3:30)

• Data visualization and reporting
  • Static visualization, e.g., Gviz
  • shiny for interactive visualizations
  • ReportingTools for effective and flexible reports
  • Sharing interactive results to your less-fortunate R-illiterate colleagues