Introduction to RNA-Seq

class: center, middle, inverse, title-slide

# Introduction to RNA-Seq
## Introduction To Bioinformatics Using NGS Data
### Dag Ahrén • 22-May-2019
### NBIS, SciLifeLab

---

exclude: true
count: false

---
name: contents
class: spaced

## Contents

* [RNA Sequencing](#intro)
* [Workflow](#workflow)
* [DGE Workflow](#workflow-dge)
* [ReadQC](#read-qc)
* [Mapping](#mapping-intro)
* [Alignment QC](#alignment-qc)
* [Quantification](#quantification-counts)
* [Normalisation](#normalisation)
* [Exploratory](#exploratory-heatmap)
* [DGE](#dge-1)
* [Functional analyses](#functional-analysis-1)
* [Summary](#summary)
* [Help](#help)

---
name: intro
class: spaced

## RNA Sequencing

![](images/rnaseq_transcription.svg)

- The transcriptome is spatially and temporally dynamic
- Data comes from functional units (coding regions)
- Only a tiny fraction of the genome

---
name: youRNASeq
class: spaced

## How many do RNASeq?

How many of you have/will have RNASeq as a component in your research?

- Raise of hands

[Menti.com](https://www.menti.com)

---
name: applications
class: spaced

## Applications

- Identify gene sequences in genomes
- Learn about gene function
- Differential gene expression
- Explore isoform and allelic expression
- Understand co-expression, pathways and networks
- Gene fusion
- RNA editing
- Phylogeny
- Gene discovery
- Other

---
name: workflow
class: spaced

## Workflow

.size-80[![](images/rnaseq_workflow.svg)]

.citation[
.cite[ Conesa, Ana, *et al.* "A survey of best practices for RNA-seq data analysis." [Genome biology 17.1 (2016): 13](https://genomebiology.biomedcentral.com/articles/10.1186/s13059-016-0881-8)]
]

---
name: exp-design
class: spaced

## Experimental design

.pull-left-50[
- Balanced design
- Technical replicates not necessary (.altcol[Marioni *et al.*, 2008])
- Biological replicates: 6 - 12 (.altcol[Schurch *et al.*, 2016])
- ENCODE consortium
- Previous publications
- Power analysis

[RnaSeqSampleSize](https://cqs.mc.vanderbilt.edu/shiny/RnaSeqSampleSize/) (Power analysis), [Scotty](http://scotty.genetics.utah.edu/) (Power analysis with cost)
]

.pull-right-50[
.size-90[![](images/batch-effect.svg)]
]

.citation[
.cite[ Busby, Michele A., *et al.* "Scotty: a web tool for designing RNA-Seq experiments to measure differential gene expression." [Bioinformatics 29.5 (2013): 656-657](https://academic.oup.com/bioinformatics/article/29/5/656/252753)] 
.cite[ Marioni, John C., *et al.* "RNA-seq: an assessment of technical reproducibility and comparison with gene expression arrays." [Genome research (2008)](https://genome.cshlp.org/content/18/9/1509.long)] 
.cite[ Schurch, Nicholas J., *et al.* "How many biological replicates are needed in an RNA-seq experiment and which differential expression tool should you use?." [Rna (2016)](http://rnajournal.cshlp.org/content/early/2016/03/30/rna.053959.115.abstract)] 
.cite[ Zhao, Shilin, *et al.* "RnaSeqSampleSize: real data based sample size estimation for RNA sequencing." [BMC bioinformatics 19.1 (2018): 191](https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-018-2191-5)] 
]

---
name: rna-extraction
class: spaced

## RNA extraction

.pull-left-50[
- Sample processing and storage
- Total RNA/mRNA/small RNA
- DNAse treatment
- Quantity & quality
- RIN values (Strong effect)
- Batch effect
- Extraction method bias (GC bias)
]

.pull-right-50[
.size-70[![](images/degradation.jpg)]
]

.citation[
.cite[ Romero, Irene Gallego, *et al*. "RNA-seq: impact of RNA degradation on transcript quantification." [BMC biology 12.1 (2014): 42](https://bmcbiol.biomedcentral.com/articles/10.1186/1741-7007-12-42)] 
.cite[ Kim, Young-Kook, *et al*. "Short structured RNAs with low GC content are selectively lost during extraction from a small number of cells." [Molecular cell 46.6 (2012): 893-895](https://www.cell.com/molecular-cell/fulltext/S1097-2765(12)00481-9).] 
]

---
name: library-prep
class: spaced

## Library prep

.pull-left-50[
- PolyA selection
- rRNA depletion
- Size selection
- PCR amplification (See section PCR duplicates)
- Stranded (directional) libraries
  - Accurately identify sense/antisense transcript
  - Resolve overlapping genes
- Exome capture
- Library normalisation
- Batch effect
]

.pull-right-50[
![](images/rnaseq_library_prep.svg)
]

.citation[
.cite[ Zhao, Shanrong, et al. "Comparison of stranded and non-stranded RNA-seq transcriptome profiling and investigation of gene overlap." [BMC genomics 16.1 (2015): 675](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4559181/)] 
.cite[ Levin, Joshua Z., et al. "Comprehensive comparative analysis of strand-specific RNA sequencing methods." [Nature methods 7.9 (2010): 709](https://www.nature.com/articles/nmeth.1491)] 
]

---
name: sequencing

## Sequencing

.pull-left-60[
- Sequencer (Illumina/PacBio)
- Read length
  - Greater than 50bp does not improve DGE
  - Longer reads better for isoforms
- Pooling samples
- Sequencing depth (Coverage/Reads per sample)
- Single-end reads (Cheaper)
- Paired-end reads
  - Increased mappable reads
  - Increased power in assemblies
  - Better for structural variation and isoforms
  - Decreased false-positives for DGE
]

.pull-right-40[
![](images/rnaseq_read_type.svg)
]

.citation[
.cite[ Chhangawala, Sagar, et al. "The impact of read length on quantification of differentially expressed genes and splice junction detection." [Genome biology 16.1 (2015): 131](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4531809/)] 
.cite[ Corley, Susan M., et al. "Differentially expressed genes from RNA-Seq and functional enrichment results are affected by the choice of single-end versus paired-end reads and stranded versus non-stranded protocols." [BMC genomics 18.1 (2017): 399](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5442695/)] 
.cite[ Liu, Yuwen, Jie Zhou, and Kevin P. White. "RNA-seq differential expression studies: more sequence or more replication?." [Bioinformatics 30.3 (2013): 301-304](https://academic.oup.com/bioinformatics/article/30/3/301/228651)]
.cite[ Comparison of PE and SE for RNA-Seq, [SciLifeLab](https://ngisweden.scilifelab.se/file/1540-1_Comparison_of_PE_and_SE_for_RNA-seq.pdf)] 
]

---
name: workflow-dge

## Workflow • DGE

![](images/rnaseq_workflow_dge.svg)

---
name: de-novo-assembly
class: spaced

## De-Novo assembly

- When no reference genome available
- To identify novel genes/transcripts/isoforms
- Identify fusion genes
- Assemble transcriptome from short reads
- Access quality of assembly and refine
- Map reads back to assembled transcriptome

[Trinity](https://github.com/trinityrnaseq/trinityrnaseq/wiki), [SOAPdenovo-Trans](http://soap.genomics.org.cn/SOAPdenovo-Trans.html), [Oases](https://www.ebi.ac.uk/~zerbino/oases/), [rnaSPAdes](http://cab.spbu.ru/software/rnaspades/)

.citation[
.cite[ Hsieh, Ping-Han *et al*., "Effect of de novo transcriptome assembly on transcript quantification" [2018 bioRxiv 380998](https://www.biorxiv.org/content/early/2018/08/22/380998)] 
.cite[ Wang, Sufang, and Michael Gribskov. "Comprehensive evaluation of de novo transcriptome assembly programs and their effects on differential gene expression analysis." [Bioinformatics 33.3 (2017): 327-333](https://academic.oup.com/bioinformatics/article/33/3/327/2580374)] 
]

---
name: read-qc

## Read QC

- Number of reads
- Per base sequence quality
- Per sequence quality score
- Per base sequence content
- Per sequence GC content
- Per base N content
- Sequence length distribution
- Sequence duplication levels
- Overrepresented sequences
- Adapter content
- Kmer content

[FastQC](https://www.bioinformatics.babraham.ac.uk/projects/fastqc/), [MultiQC](http://multiqc.info/)

https://sequencing.qcfail.com/

![](images/qcfail.jpg)

---
name: fastqc

## FastQC

.size-70[[![](images/fastqc_good.png)](http://www.bioinformatics.babraham.ac.uk/projects/fastqc/good_sequence_short_fastqc.html)]

.size-70[[![](images/fastqc_bad.png)](http://www.bioinformatics.babraham.ac.uk/projects/fastqc/bad_sequence_fastqc.html)]

---
name: read-qc-2

## Read QC • PBSQ, PSQS

.size-85[.small[**Per base sequence quality**] ![](images/pbsq.jpg)]
.size-85[.small[**Per sequence quality scores**] ![](images/psqs.jpg)]

---
name: read-qc-3

## Read QC • PBSC, PSGC

.size-85[.small[**Per base sequence content**] ![](images/pbsc.jpg)]
.size-85[.small[**Per sequence GC content**] ![](images/psgc.jpg)]

---
name: read-qc-4

## Read QC • SDL, AC

.size-85[.small[**Sequence duplication level**] ![](images/sdl.jpg)]
.size-85[.small[**Adapter content**] ![](images/ac.jpg)]

---
name: trimming

## Trimming

.pull-left-50[
- Trim IF necessary
  - Synthetic bases can be an issue for SNP calling
  - Insert size distribution may be more important for assemblers
- Trim/Clip/Filter reads
- Remove adapter sequences
- Trim reads by quality
- Sliding window trimming
- Filter by min/max read length
  - Remove reads less than ~18nt
- Demultiplexing/Splitting

[Cutadapt](https://github.com/marcelm/cutadapt/), [fastp](https://github.com/OpenGene/fastp), [Skewer](https://github.com/relipmoc/skewer), [Prinseq](http://prinseq.sourceforge.net/)
]

.pull-right-50[
.size-90[![](images/rnaseq_read_through.svg)]
]

---
name: mapping-intro

## Mapping

![](images/rnaseq_mapping.svg)

- Aligning reads back to a reference sequence
- Mapping to genome vs transcriptome
- Splice-aware alignment (genome)

[STAR](https://github.com/alexdobin/STAR), [HiSat2](https://ccb.jhu.edu/software/hisat2/index.shtml), [GSNAP](http://research-pub.gene.com/gmap/), [Novoalign](http://www.novocraft.com/products/novoalign/) (Commercial)

.citation[
.cite[ Baruzzo, Giacomo, *et al*. "Simulation-based comprehensive benchmarking of RNA-seq aligners." [Nature methods 14.2 (2017): 135](https://www.nature.com/articles/nmeth.4106)]
]

---
name: aligner-speed

## Aligners • Speed

.pull-left-50[
![](images/compare-aligners-speed.jpg)
]

.pull-right-50[
|Program|Time_Min|Memory_GB|
|-------|--------|---------|
|HISATx1|22.7|4.3|
|HISATx2|47.7|4.3|
|HISAT|26.7|4.3|
|STAR|25|28|
|STARx2|50.5|28|
|GSNAP|291.9|20.2|
|TopHat2|1170|4.3|
]

---
name: aligner-accuracy

## Aligners • Accuracy

.size-70[![](images/compare-aligners-quality.jpg)]

.size-40[![](images/accuracy.svg)]

---
name: mapping-required

## Mapping

- Reads (FASTQ)

```
@ST-E00274:179:HHYMLALXX:8:1101:1641:1309 1:N:0:NGATGT
NCATCGTGGTATTTGCACATCTTTTCTTATCAAATAAAAAGTTTAACCTACTCAGTTATGCGCATACGTTTTTTGATGGCATTTCCATAAACCGATTTTTTTTTTATGCACGTACCCAAAACGTGCAGAAAAATACGCTGCTAGAAATGTA
+
#AAAFAFA<-AFFJJJAFA-FFJJJJFFFAJJJJ-<FFJJJ-A-F-7--FA7F7-----FFFJFA<FFFFJ<AJ--FF-A<A-<JJ-7-7-<FF-FFFJAFFAA--A--7FJ-7----77-A--7F7)---7F-A----7)7-----7<<-
```

`@instrument:runid:flowcellid:lane:tile:xpos:ypos read:isfiltered:controlnumber:sampleid`

- Reference Genome/Transcriptome (FASTA)

```
>1 dna:chromosome chromosome:GRCz10:1:1:58871917:1 REF
GATCTTAAACATTTATTCCCCCTGCAAACATTTTCAATCATTACATTGTCATTTCCCCTC
CAAATTAAATTTAGCCAGAGGCGCACAACATACGACCTCTAAAAAAGGTGCTGTAACATG
```

- Annotation (GTF/GFF)

```
#!genome-build GRCz10
#!genebuild-last-updated 2016-11
4       ensembl_havana  gene    6732    52059   .       -       .       gene_id "ENSDARG00000104632"; gene_version "2"; gene_name "rerg"; gene_source "ensembl_havana"; gene_biotype "protein_coding"; havana_gene "OTTDARG00000044080"; havana_gene_version "1";
```

`seq source feature start end score strand frame attribute`

.citation[
.cite[ Illumina read name [format](http://support.illumina.com/content/dam/illumina-support/help/BaseSpaceHelp_v2/Content/Vault/Informatics/Sequencing_Analysis/BS/swSEQ_mBS_FASTQFiles.htm), GTF [format](https://www.ensembl.org/info/website/upload/gff.html)] 
]

---
name: alignment

## Alignment

- SAM/BAM (Sequence Alignment Map format)

```
ST-E00274:188:H3JWNCCXY:4:1102:32431:49900 163 1 1 60 8S139M4S = 385 535 TATTTAGAGATCTTAAACATCCATTCCCCCTGCAAACATTTTCAATCATTACATTGTCATTTTCCCTCCAAATTAAATTTAGCCAGAGGCGCACAACATACGACCTCTAAAAAAGGTGCTGGAACATGTACCTATATGCAGCACCACCATC AAAFAFFAFFFFJ7FFFFJ<JAFA7F-<AJ7JJ<FFFJ--<FAJF<7<7FAFJ-<AFA<-JJJ-AF-AJ-FF<F--A<FF<-7777-7JA-77A---F-7AAFF-FJA--77FJ<--77)))7<JJA<J77<-------<7--))7)))7- NM:i:4 MD:Z:12T0T40C58T25 AS:i:119 XS:i:102 XA:Z:17,-53287490,4S33M4D114M,11; MQ:i:60 MC:Z:151M RG:Z:ST-E00274_188_H3JWNCCXY_4
```

`query flag ref pos mapq cigar mrnm mpos tlen seq qual opt`

|Format|Size_GB|
|------|-------|
|SAM|7.4|
|BAM|1.9|
|CRAM lossless Q|1.4|
|CRAM 8 bins Q|0.8|
|CRAM no Q|0.26|

.citation[
.cite[ SAM file [format](http://www.htslib.org/doc/sam.html)]
]

---
name: vis-tview

## Visualisation • `tview`

`samtools tview alignment.bam genome.fasta`

![](images/tview.png)

---
name: vis-igv

## Visualisation • IGV

![](images/igv.png)

[IGV](http://software.broadinstitute.org/software/igv/), [UCSC Genome Browser](https://genome.ucsc.edu/)

---
name: vis-seqmonk

## Visualisation • SeqMonk

![](images/seqmonk.jpg)

[SeqMonk](https://www.bioinformatics.babraham.ac.uk/projects/seqmonk/)

---
name: alignment-qc
class: spaced

## Alignment QC

- Number of reads mapped/unmapped/paired etc
- Uniquely mapped
- Insert size distribution
- Coverage
- Gene body coverage
- Biotype counts / Chromosome counts
- Counts by region: gene/intron/non-genic
- Sequencing saturation
- Strand specificity

STAR (final log file), samtools > stats, bamtools > stats, [QoRTs](https://hartleys.github.io/QoRTs/), [RSeQC](http://rseqc.sourceforge.net/), [Qualimap](http://qualimap.bioinfo.cipf.es/)

---
name: star-log

## Alignment QC • STAR Log

MultiQC can be used to summarise and plot STAR log files.

.size-95[![](images/star_alignment_plot.svg)]

---
name: qorts-region

## Alignment QC • Features

QoRTs was run on all samples and summarised using MultiQC.

.size-95[![](images/qorts_alignments.svg)]

---
name: qorts-plots

## QoRTs

.size-95[![](images/qorts.png)]

---
name: alignment-qc-1

## Alignment QC

.pull-left-50[
**Soft clipping**
![](images/clipping_good.png)
]

.pull-right-50[
**Gene body coverage**
![](images/gene-body-coverage.png)
]

---
name: alignment-qc-2

## Alignment QC

.pull-left-50[
**Insert size**
![](images/inner_distance.png)
]

.pull-right-50[
**Saturation curve**
![](images/saturation.jpg)
]

---
name: quantification-counts
class: spaced

## Quantification • Counts

.pull-left-50[
- Read counts = gene expression
- Reads can be quantified on any feature (gene, transcript, exon etc)
- Intersection on gene models
- Gene/Transcript level

![](images/rnaseq_counts.svg)

[featureCounts](http://bioinf.wehi.edu.au/featureCounts/), [HTSeq](https://github.com/simon-anders/htseq)
]

.pull-right-50[
.size-85[.center[![](images/rnaseq_union.svg)]]
]

---
name: quant-pcr-multi

## Quantification

.pull-left-50[
**PCR duplicates**

- Ignore for RNA-Seq data
- Computational deduplication (Don't!)
- Use PCR-free library-prep kits
- Use UMIs during library-prep

**Multi-mapping**

- Added (BEDTools multicov)
- Discard (featureCounts, HTSeq)
- Distribute counts (Cufflinks)
- Rescue
  - Probabilistic assignment (Rcount, Cufflinks)
  - Prioritise features (Rcount)
  - Probabilistic assignment with EM (RSEM)
]

.pull-right-50[
![](images/pcr-duplicates.png)
]

.citation[
.cite[ Fu, Yu, *et al*. "Elimination of PCR duplicates in RNA-seq and small RNA-seq using unique molecular identifiers." [BMC genomics 19.1 (2018): 531](https://bmcgenomics.biomedcentral.com/articles/10.1186/s12864-018-4933-1)] 
.cite[ Parekh, Swati, *et al*. "The impact of amplification on differential expression analyses by RNA-seq." [Scientific reports 6 (2016): 25533](https://www.nature.com/articles/srep25533)] 
.cite[ Klepikova, Anna V., *et al*. "Effect of method of deduplication on estimation of differential gene expression using RNA-seq." [PeerJ 5 (2017): e3091](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5357343/)] 
]

---
name: quantification-abundance

## Quantification • Abundance

- Count methods
 - Provide no inference on isoforms
 - Cannot accurately measure fold change

- Probabilistic assignment
 - Deconvolute ambiguous mappings
 - Transcript-level
 - cDNA reference

**Kallisto, Salmon**

- Ultra-fast & alignment-free
- Subsampling & quantification confidence
- Transcript-level estimates improves gene-level estimates
- Kallisto/Salmon > transcript-counts > `tximport()` > gene-counts

[RSEM](https://deweylab.github.io/RSEM/), [Kallisto](https://pachterlab.github.io/kallisto/), [Salmon](https://combine-lab.github.io/salmon/), [Cufflinks2](http://cole-trapnell-lab.github.io/cufflinks/)

.citation[
.cite[ Soneson, Charlotte, *et al*. "Differential analyses for RNA-seq: transcript-level estimates improve gene-level inferences." [F1000Research 4 (2015)](https://f1000research.com/articles/4-1521/v2)] 
.cite[ Zhang, Chi, *et al*. "Evaluation and comparison of computational tools for RNA-seq isoform quantification." [BMC genomics 18.1 (2017): 583](https://bmcgenomics.biomedcentral.com/articles/10.1186/s12864-017-4002-1)] 
]

---
name: quantification-qc

## Quantification QC

```
ENSG00000000003    140   242   188   143   287   344   438   280   253
ENSG00000000005    0     0     0     0     0     0     0     0     0
ENSG00000000419    69    98    77    55    52    94    116   79    69
ENSG00000000457    56    75    104   79    157   205   183   178   153
ENSG00000000460    33    27    23    19    27    42    69    44    40
ENSG00000000938    7     38    13    17    35    76    53    37    24
ENSG00000000971    545   878   694   636   647   216   492   798   323
ENSG00000001036    79    154   74    80    128   167   220   147   72
```

.pull-left-50[
- Pairwise correlation between samples must be high (>0.9)

.size-60[![](images/correlation.png)]

]

.pull-right-50[
- Count QC using RNASeqComp

.size-80[![](images/rnaseqcomp.gif)]

]

[RNASeqComp](https://bioconductor.org/packages/release/bioc/html/rnaseqcomp.html)

.citation[
.cite[ Teng, Mingxiang, *et al*. "A benchmark for RNA-seq quantification pipelines." [Genome biology 17.1 (2016): 74](https://genomebiology.biomedcentral.com/articles/10.1186/s13059-016-0940-1)] 
]

---
name: multiqc

## MultiQC

[![](images/multiqc.png)](https://multiqc.info/examples/rna-seq/multiqc_report.html)

---
name: normalisation

## Normalisation

- Control for Sequencing depth & compositional bias
- Median of Ratios (DESeq2) and TMM (edgeR) perform the best

![](images/normalisation.png)

- For DGE using DGE packages, use raw counts
- For clustering, heatmaps etc use VST, VOOM or RLOG
- For own analysis, plots etc, use TPM
- Other solutions: spike-ins/house-keeping genes

.citation[
.cite[ Dillies, Marie-Agnes, *et al*. "A comprehensive evaluation of normalization methods for Illumina high-throughput RNA sequencing data analysis." [Briefings in bioinformatics 14.6 (2013): 671-683](https://www.ncbi.nlm.nih.gov/pubmed/22988256)] 
.cite[ Evans, Ciaran, Johanna Hardin, and Daniel M. Stoebel. "Selecting between-sample RNA-Seq normalization methods from the perspective of their assumptions." [Briefings in bioinformatics (2017)](https://arxiv.org/abs/1609.00959)] 
.cite[ Wagner, Gunter P., Koryu Kin, and Vincent J. Lynch. "Measurement of mRNA abundance using RNA-seq data: RPKM measure is inconsistent among samples." [Theory in biosciences 131.4 (2012): 281-285](https://link.springer.com/article/10.1007/s12064-012-0162-3)] 
]

---
name: exploratory-heatmap

## Exploratory • Heatmap

- Remove lowly expressed genes
- Transform raw counts to VST, VOOM, RLOG, TPM etc
- Sample-sample clustering heatmap

[`pheatmap()`](https://github.com/raivokolde/pheatmap)

---
name: exploratory-mds

## Exploratory • MDS

<div id="htmlwidget-cc80a985b57a0099bf80" style="width:750px;height:500px;" class="plotly html-widget"></div>
<script type="application/json" data-for="htmlwidget-cc80a985b57a0099bf80">{"x":{"visdat":{"44f2dac2bba":["function () ","plotlyVisDat"]},"cur_data":"44f2dac2bba","attrs":{"44f2dac2bba":{"x":{},"y":{},"z":{},"text":{},"hoverinfo":"text","colors":["#A6CEE3","#1F78B4","#B2DF8A","#33A02C","#FB9A99","#E31A1C","#FDBF6F","#FF7F00"],"alpha_stroke":1,"sizes":[10,100],"spans":[1,20],"type":"scatter3d","mode":"markers","color":{},"inherit":true}},"layout":{"width":750,"height":500,"margin":{"b":40,"l":60,"t":25,"r":10},"scene":{"xaxis":{"title":"V1"},"yaxis":{"title":"V2"},"zaxis":{"title":"V3"}},"hovermode":"closest","showlegend":true},"source":"A","config":{"modeBarButtonsToAdd":[{"name":"Collaborate","icon":{"width":1000,"ascent":500,"descent":-50,"path":"M487 375c7-10 9-23 5-36l-79-259c-3-12-11-23-22-31-11-8-22-12-35-12l-263 0c-15 0-29 5-43 15-13 10-23 23-28 37-5 13-5 25-1 37 0 0 0 3 1 7 1 5 1 8 1 11 0 2 0 4-1 6 0 3-1 5-1 6 1 2 2 4 3 6 1 2 2 4 4 6 2 3 4 5 5 7 5 7 9 16 13 26 4 10 7 19 9 26 0 2 0 5 0 9-1 4-1 6 0 8 0 2 2 5 4 8 3 3 5 5 5 7 4 6 8 15 12 26 4 11 7 19 7 26 1 1 0 4 0 9-1 4-1 7 0 8 1 2 3 5 6 8 4 4 6 6 6 7 4 5 8 13 13 24 4 11 7 20 7 28 1 1 0 4 0 7-1 3-1 6-1 7 0 2 1 4 3 6 1 1 3 4 5 6 2 3 3 5 5 6 1 2 3 5 4 9 2 3 3 7 5 10 1 3 2 6 4 10 2 4 4 7 6 9 2 3 4 5 7 7 3 2 7 3 11 3 3 0 8 0 13-1l0-1c7 2 12 2 14 2l218 0c14 0 25-5 32-16 8-10 10-23 6-37l-79-259c-7-22-13-37-20-43-7-7-19-10-37-10l-248 0c-5 0-9-2-11-5-2-3-2-7 0-12 4-13 18-20 41-20l264 0c5 0 10 2 16 5 5 3 8 6 10 11l85 282c2 5 2 10 2 17 7-3 13-7 17-13z m-304 0c-1-3-1-5 0-7 1-1 3-2 6-2l174 0c2 0 4 1 7 2 2 2 4 4 5 7l6 18c0 3 0 5-1 7-1 1-3 2-6 2l-173 0c-3 0-5-1-8-2-2-2-4-4-4-7z m-24-73c-1-3-1-5 0-7 2-2 3-2 6-2l174 0c2 0 5 0 7 2 3 2 4 4 5 7l6 18c1 2 0 5-1 6-1 2-3 3-5 3l-174 0c-3 0-5-1-7-3-3-1-4-4-5-6z"},"click":"function(gd) { \n // is this being viewed in RStudio?\n if (location.search == '?viewer_pane=1') {\n alert('To learn about plotly for collaboration, visit:\\n https://cpsievert.github.io/plotly_book/plot-ly-for-collaboration.html');\n } else {\n window.open('https://cpsievert.github.io/plotly_book/plot-ly-for-collaboration.html', '_blank');\n }\n }"}],"cloud":false},"data":[{"x":[-5.33185049177959,-7.92209817002156,-8.77692340480147],"y":[0.0432736817716184,-1.26280672685875,-1.86761681936745],"z":[1.04060988518825,-1.41087805672964,-2.80687594167858],"text":["ID: <\/b>121T10571_121 NGS ID: <\/b>P8304_117","ID: <\/b>121T10571_123 NGS ID: <\/b>P8304_119","ID: <\/b>121T10571_124 NGS ID: <\/b>P8304_120"],"hoverinfo":["text","text","text"],"type":"scatter3d","mode":"markers","name":"121T10571_12","marker":{"color":"rgba(166,206,227,1)","line":{"color":"rgba(166,206,227,1)"}},"textfont":{"color":"rgba(166,206,227,1)"},"error_y":{"color":"rgba(166,206,227,1)"},"error_x":{"color":"rgba(166,206,227,1)"},"line":{"color":"rgba(166,206,227,1)"},"frame":null},{"x":[-5.73550373147591,-6.21011122708704,-7.35383060771142,-6.59073194911667],"y":[-6.23950665019801,-6.50430598496082,-6.51161261887537,-5.34119076269499],"z":[7.24464218182632,7.38236724522226,6.0494637371315,5.38523664148418],"text":["ID: <\/b>134_T6443_11_1 NGS ID: <\/b>P8304_113","ID: <\/b>134_T6443_11_2 NGS ID: <\/b>P8304_114","ID: <\/b>134_T6443_11_3 NGS ID: <\/b>P8304_115","ID: <\/b>134_T6443_11_4 NGS ID: <\/b>P8304_116"],"hoverinfo":["text","text","text","text"],"type":"scatter3d","mode":"markers","name":"134_T6443_11","marker":{"color":"rgba(31,120,180,1)","line":{"color":"rgba(31,120,180,1)"}},"textfont":{"color":"rgba(31,120,180,1)"},"error_y":{"color":"rgba(31,120,180,1)"},"error_x":{"color":"rgba(31,120,180,1)"},"line":{"color":"rgba(31,120,180,1)"},"frame":null},{"x":[-4.6497016412439,-5.10088962658455,-6.00012520521127,-4.41802453955813],"y":[3.09565456848034,1.18448065666745,2.63713128689337,0.792488532872201],"z":[1.84057242738463,1.06376484620365,0.0898534016981559,3.4308686704621],"text":["ID: <\/b>153_ST132_131 NGS ID: <\/b>P8304_101","ID: <\/b>153_ST132_132 NGS ID: <\/b>P8304_102","ID: <\/b>153_ST132_133 NGS ID: <\/b>P8304_103","ID: <\/b>153_ST132_134 NGS ID: <\/b>P8304_104"],"hoverinfo":["text","text","text","text"],"type":"scatter3d","mode":"markers","name":"153_ST132_13","marker":{"color":"rgba(178,223,138,1)","line":{"color":"rgba(178,223,138,1)"}},"textfont":{"color":"rgba(178,223,138,1)"},"error_y":{"color":"rgba(178,223,138,1)"},"error_x":{"color":"rgba(178,223,138,1)"},"line":{"color":"rgba(178,223,138,1)"},"frame":null},{"x":[-5.88085854951221,-7.76790916625506,-6.25504507611664,-8.3764530117268],"y":[2.28585250238002,1.36384290506196,2.03080210880601,0.732589724202596],"z":[2.53910291233844,-0.247836766653928,1.897953832452,-0.233518938722922],"text":["ID: <\/b>24_TD9169_081 NGS ID: <\/b>P8304_109","ID: <\/b>24_TD9169_082 NGS ID: <\/b>P8304_110","ID: <\/b>24_TD9169_083 NGS ID: <\/b>P8304_111","ID: <\/b>24_TD9169_084 NGS ID: <\/b>P8304_112"],"hoverinfo":["text","text","text","text"],"type":"scatter3d","mode":"markers","name":"24_TD9169_08","marker":{"color":"rgba(51,160,44,1)","line":{"color":"rgba(51,160,44,1)"}},"textfont":{"color":"rgba(51,160,44,1)"},"error_y":{"color":"rgba(51,160,44,1)"},"error_x":{"color":"rgba(51,160,44,1)"},"line":{"color":"rgba(51,160,44,1)"},"frame":null},{"x":[-3.88932422329194,-0.0729461302680156],"y":[-1.69996167204044,1.77387505239342],"z":[1.96344736048856,5.74550696017447],"text":["ID: <\/b>29_T1942_083 NGS ID: <\/b>P8304_123","ID: <\/b>29_T1942_084 NGS ID: <\/b>P8304_124"],"hoverinfo":["text","text"],"type":"scatter3d","mode":"markers","name":"29_T1942_08","marker":{"color":"rgba(251,154,153,1)","line":{"color":"rgba(251,154,153,1)"}},"textfont":{"color":"rgba(251,154,153,1)"},"error_y":{"color":"rgba(251,154,153,1)"},"error_x":{"color":"rgba(251,154,153,1)"},"line":{"color":"rgba(251,154,153,1)"},"frame":null},{"x":[-10.2163172144276,-10.596078285728,-10.2353954116357,-9.33007941422633],"y":[-0.490190092826988,-2.26662762461056,-0.333539418783384,-0.0554168358930101],"z":[-8.46229337166764,-8.5409366265751,-8.94003110723548,-8.17218218493106],"text":["ID: <\/b>61_T1538_071 NGS ID: <\/b>P8304_105","ID: <\/b>61_T1538_072 NGS ID: <\/b>P8304_106","ID: <\/b>61_T1538_073 NGS ID: <\/b>P8304_107","ID: <\/b>61_T1538_074 NGS ID: <\/b>P8304_108"],"hoverinfo":["text","text","text","text"],"type":"scatter3d","mode":"markers","name":"61_T1538_07","marker":{"color":"rgba(227,26,28,1)","line":{"color":"rgba(227,26,28,1)"}},"textfont":{"color":"rgba(227,26,28,1)"},"error_y":{"color":"rgba(227,26,28,1)"},"error_x":{"color":"rgba(227,26,28,1)"},"line":{"color":"rgba(227,26,28,1)"},"frame":null},{"x":[21.8493946916054,22.7364624888557,22.8684179809813,23.401941446664],"y":[-5.14341595704935,-6.83851698747042,-14.3376040568602,-8.02483675387452],"z":[1.5998970735274,0.863612874314833,-7.37316041937378,-1.33865318696138],"text":["ID: <\/b>TD11549_17_O1 NGS ID: <\/b>P8304_125","ID: <\/b>TD11549_17_O2 NGS ID: <\/b>P8304_126","ID: <\/b>TD11549_17_O3 NGS ID: <\/b>P8304_127","ID: <\/b>TD11549_17_O4 NGS ID: <\/b>P8304_128"],"hoverinfo":["text","text","text","text"],"type":"scatter3d","mode":"markers","name":"TD11549_17_O","marker":{"color":"rgba(253,191,111,1)","line":{"color":"rgba(253,191,111,1)"}},"textfont":{"color":"rgba(253,191,111,1)"},"error_y":{"color":"rgba(253,191,111,1)"},"error_x":{"color":"rgba(253,191,111,1)"},"line":{"color":"rgba(253,191,111,1)"},"frame":null},{"x":[14.4253959187119,13.2889594437547,10.7843904183293,11.3552346888775],"y":[15.1368414992242,12.1873998562284,13.2999768462921,10.3529397410905],"z":[1.81496155659233,-0.3728142071169,-0.865737731812759,-1.18694306702993],"text":["ID: <\/b>TD11558_17_L1 NGS ID: <\/b>P8304_129","ID: <\/b>TD11558_17_L2 NGS ID: <\/b>P8304_130","ID: <\/b>TD11558_17_L3 NGS ID: <\/b>P8304_131","ID: <\/b>TD11558_17_L4 NGS ID: <\/b>P8304_132"],"hoverinfo":["text","text","text","text"],"type":"scatter3d","mode":"markers","name":"TD11558_17_L","marker":{"color":"rgba(255,127,0,1)","line":{"color":"rgba(255,127,0,1)"}},"textfont":{"color":"rgba(255,127,0,1)"},"error_y":{"color":"rgba(255,127,0,1)"},"error_x":{"color":"rgba(255,127,0,1)"},"line":{"color":"rgba(255,127,0,1)"},"frame":null}],"highlight":{"on":"plotly_click","persistent":false,"dynamic":false,"selectize":false,"opacityDim":0.2,"selected":{"opacity":1},"debounce":0},"base_url":"https://plot.ly"},"evals":["config.modeBarButtonsToAdd.0.click"],"jsHooks":[]}</script>

[`cmdscale()`](https://stat.ethz.ch/R-manual/R-devel/library/stats/html/cmdscale.html), [plotly](https://plot.ly/r/)

---
name: batch-correction
class: spaced

## Batch correction

- Estimate variation explained by variables (PVCA)

.size-70[![](images/pvca.png)]

- Find confounding effects as surrogate variables (SVA)
- Model known batches in the LM/GLM model
- Correct known batches (ComBat)(Harsh!)
- Interactively evaluate batch effects and correction (BatchQC)

[SVA](http://bioconductor.org/packages/release/bioc/html/sva.html), [PVCA](https://bioconductor.org/packages/release/bioc/html/pvca.html), [BatchQC](http://bioconductor.org/packages/release/bioc/html/BatchQC.html)

.citation[
.cite[ Liu, Qian, and Marianthi Markatou. "Evaluation of methods in removing batch effects on RNA-seq data." [Infectious Diseases and Translational Medicine 2.1 (2016): 3-9](http://www.tran-med.com/article/2016/2411-2917-2-1-3.html)] 
.cite[ Manimaran, Solaiappan, et al. "BatchQC: interactive software for evaluating sample and batch effects in genomic data." [Bioinformatics 32.24 (2016): 3836-3838](https://academic.oup.com/bioinformatics/article/32/24/3836/2525651)] 
]

---
name: dge-1

## DGE

- DESeq2, edgeR (Neg-binom > GLM > Test), Limma-Voom (Neg-binom > Voom-transform > LM > Test)
- DESeq2 `~age+condition`
  - Estimate size factors `estimateSizeFactors()`
  - Estimate gene-wise dispersion `estimateDispersions()`
  - Fit curve to gene-wise dispersion estimates
  - Shrink gene-wise dispersion estimates
  - GLM fit for each gene
  - Wald test `nbinomWaldTest()`

[DESeq2](), [edgeR](), [Limma-Voom]()

.citation[
.cite[ Seyednasrollah, Fatemeh, *et al*. "Comparison of software packages for detecting differential expression in RNA-seq studies." [Briefings in bioinformatics 16.1 (2013): 59-70](https://academic.oup.com/bib/article/16/1/59/240754)] 
]

---
name: dge-2

## DGE

- Results `results()`

```
## log2 fold change (MLE): type type2 vs control 
## Wald test p-value: type type2 vs control 
## DataFrame with 1 row and 6 columns
## baseMean log2FoldChange lfcSE
## <numeric> <numeric> <numeric>
## ENSG00000000003 242.307796723287 -0.93292608960856 0.11428515031257
## stat pvalue
## <numeric> <numeric>
## ENSG00000000003 -8.16314356727017 3.26416150297406e-16
## padj
## <numeric>
## ENSG00000000003 1.36240610021329e-14
```

- Summary `summary()`

```
## 
## out of 17889 with nonzero total read count
## adjusted p-value < 0.1
## LFC > 0 (up) : 4526, 25%
## LFC < 0 (down) : 5062, 28%
## outliers [1] : 25, 0.14%
## low counts [2] : 0, 0%
## (mean count < 3)
## [1] see 'cooksCutoff' argument of ?results
## [2] see 'independentFiltering' argument of ?results
```

---
name: dge-3

## DGE

.pull-left-50[

- MA plot `plotMA()`

- Volcano plot

]

.pull-right-50[

- Normalised counts `plotCounts()`

<img src="images/scattered.gif" style="height:220px; position:fixed; right:0px; bottom:0px; margin-right: 130px; margin-bottom: 10px;">
]

---
name: functional-analysis-1

## Functional analysis • GO

- Gene enrichment analysis
- Gene set enrichment analysis (GSEA)
- Gene ontology / Reactome databases

.size-70[![](images/go.jpg)]

---
name: functional-analysis-2

## Functional analysis • Kegg

- Pathway analysis (Kegg)

![](images/kegg.png)

[DAVID](https://david.ncifcrf.gov/), [clusterProfiler](https://bioconductor.org/packages/release/bioc/html/clusterProfiler.html), [ClueGO](http://apps.cytoscape.org/apps/cluego), [ErmineJ](https://erminej.msl.ubc.ca/), [pathview](https://bioconductor.org/packages/release/bioc/html/pathview.html)

---
name: summary
class: spaced

## Summary

- Sound experimental design to avoid confounding
- Plan carefully about lib prep, sequencing etc based on experimental objective
- Biological replicates may be more important than paired-end reads or long reads
- Discard low quality bases, reads, genes and samples
- Verify that tools and methods align with data assumptions
- Experiment with multiple pipelines and tools
- QC! QC everything at every step

.large[ Conesa, Ana, *et al.* "A survey of best practices for RNA-seq data analysis." [Genome biology 17.1 (2016): 13](https://genomebiology.biomedcentral.com/articles/10.1186/s13059-016-0881-8)]

---
name: help
class: spaced

## Further learning

- Griffith lab [RNA-Seq using HiSat & StringTie tutorial](https://github.com/griffithlab/rnaseq_tutorial/wiki)
- SciLifeLab [courses](https://www.scilifelab.se/education/courses%26training)
- HBC Training [DGE using DeSeq2 tutorial](https://github.com/hbctraining/Intro-to-R-with-DGE)
- Hemberg lab [scRNA-Seq tutorial](http://hemberg-lab.github.io/scRNA.seq.course/index.html)
- [RNA-Seq Blog](https://www.rna-seqblog.com/)

---
name: end_slide
class: end-slide, middle
count: false

# Thank you. Questions?
**Also: Thanks to Roy Francis for the presentation**

.end-text[
R version 3.5.2 (2018-12-20) Platform: x86_64-apple-darwin15.6.0 (64-bit)OS: macOS High Sierra 10.13.6

<hr>

Built on : 22-May-2019 at 23:26:03

2019 • [SciLifeLab](https://www.scilifelab.se/) • [NBIS](https://nbis.se/)
]

---
name: lab1
class: spaced

## Hands-On tutorial

**Main exercise**

- 01 Check the quality of the raw reads with **FastQC**
- 02 Map the reads to the reference genome using **Star**
- 03 Assess the post-alignment quality using **QualiMap**
- 04 Count the reads overlapping with genes using **featureCounts**
- 05 Find DE genes using **edgeR** in R

**Bonus exercises**

- 01 Functional annotation of DE genes using **GO/Reactome/Kegg** databases
- 02 Visualisation of RNA-seq BAM files using **IGV** genome browser
- 03 RNA-Seq figures and plots using **R**
- 04 De-novo transcriptome assembly using **Trinity**

.large[**Data:** `/sw/courses/ngsintro/rnaseq/`] 
.large[**Work:** `/proj/g2019007/nobackup/<user>/rnaseq/`]

---
name: lab2

## Hands-On tutorial

.pull-left-50[

- Course data directory

.large[`/sw/courses/ngsintro/rnaseq/`]

.pull-right-50[

- Your work directory

.large[`/proj/g2019007/nobackup/[user]/`]

.large[
```
[user]/
rnaseq/
  +-- 1_raw/
  +-- 2_fastqc/
  +-- 3_mapping/
  +-- 4_qualimap/
  +-- 5_dge/
  +-- 6_multiqc/
  +-- reference/
  |   +-- mouse/
  |   +-- mouse_chr11/
  +-- scripts/
  +-- funannot/
  +-- assembly/
```
]
]