introduction to ngs analyses - forth-imbb · introduction to ngs analyses giorgio l papadopoulos...

Introduction to NGS analyses

Giorgio L Papadopoulos

Institute of Molecular Biology and Biotechnology

Bioinformatics Support Group

04/12/2015

Papadopoulos GL (IMBB, FORTH) IMBB NGS Seminar 04/12/2015 1 / 28

Overview

1 IntroductionDNA sequencingApplications

2 Primary AnalysisChIPseqRNAseq

3 InterpretationVisualizationAnalysis SoftwareDownstreamAnalysis

4 Online ResourcesENCODEGEOENA


Introduction DNA sequencing

Fred SangerLate 1970s

... Several million times ...Late 2000s


Introduction Applications

Reuter et al, Mol Cell, 2015


Introduction ChIPseq

Genome Wide Occupancy Profiling

Overlapping reads come from different cells!!!

NGS is a cell population readout ( 50 million cells)



Primary Sequencing Data



Genome wide, highly accurate representation of genomic statesFairly unbiased

Complementary and overlapping information


Primary Analysis ChIPseq

Raw Data

Aligned Data

Enriched

RegionsRead Density

VisualizationTarget Genes

Functional

Analysis

Distribution

Motifs

Overlaps



Raw Data

Aligned Data

Enriched

RegionsRead Density

Visualization

Target Genes

Functional

Analysis

Distribution

Motifs

Overlaps



Raw Data

Aligned Data

Enriched

RegionsRead Density

VisualizationTarget Genes

Functional

Analysis

Distribution

Motifs

Overlaps



Raw Data



Raw Data

Operations:DeMultiplexingQuality Control



Raw Data


FastQC



Raw Data


FastQC

Align Data...



Align Data...

Aligner (Mapper)

Reference GenomeHuman (hg18, hg19)Mouse (mm9, mm10)

...(iGenomes, UCSC, NCBI, EBI...)

Different mappers will usedifferent Index Formats...

bowtie.2

% of alignment...

Low percentage?Contaminations,

adapter/barcode trimming,wrong reference...

Bad Library...



Align Data...

Aligner (Mapper)

http://www.ebi.ac.uk/~nf/hts_mappers/



Different mappers will use different Index Formats...

bowtie.2

% of alignment...



Bad Library...


http://www.ebi.ac.uk/~nf/hts_mappers/


Align Data...

Aligner (Mapper)




bowtie.2

% of alignment...



Bad Library...



Align Data...

Aligner (Mapper)




bowtie.2

Fast...Versatile...

Documented and supported...

% of alignment...



Bad Library...



Align Data...

Aligner (Mapper)




bowtie.2

% of alignment...



Bad Library...



Aligned Data

SAM ⇔ BAM ⇒ BED⇓

Enriched

Regions

⇓

⇓Chr Start Stop ID Score

SET ofGENOMIC COORDINATES

(PEAKS...)

How Many???

Where???

With Who???

Function???

HOW???



Aligned Data

SAM ⇔ BAM ⇒ BED

⇓Enriched

Regions

⇓



(PEAKS...)

How Many???

Where???

With Who???

Function???

HOW???



Aligned Data

SAM ⇔ BAM ⇒ BED⇓

Enriched

Regions

⇓



(PEAKS...)

How Many???

Where???

With Who???

Function???

HOW???


Primary Analysis Peak Calling

“Peak calling is a computational method used to identify areas in agenome that have been enriched with aligned reads“

Challenge: Not one single ’peak shape’...

Number of peaksPeak distribution

Peak heightRange of measurements

Experimental variation (IP...)



Solutions:Different Peak CallersOptimal Parameters

Correct Experimental DesignGood ChIP...





Peaks represent only a ’summary’ of a ChIPseq experiment...






Useful metrics:FRiP: Fraction of Reads in Peaks (>1%)

IDR: Irreproducible Discovery RateSignificance: pValue, ChIPtoINPUT ratio, # of reads...






Empirical:Known Motif: Enriched in peaks, Central Position

Distribution: Near TSS, Within GeneBody, EnhancersTarget Genes: Gene expression changes, Functional Annotation






Empirical:Known Motif: Enriched in peaks, Central Position

Distribution: Near TSS, Within GeneBody, EnhancersTarget Genes: Gene expression changes, Functional Annotation

Browser inspection and previously known sites



TFs: MACSHMs: SICER



TFs: MACSHMs: SICER

Running MACS: macs14 -t ChIP -c Input -g mm -n Name



TFs: MACSHMs: SICER


−600 −400 −200 0 200 400 600

0.0

0.1

0.2

0.3

0.4

Peak Model

Distance to the middle

Per

cent

age

forward tagsreverse tagsshifted tags

d=119



TFs: MACSHMs: SICER


−600 −400 −200 0 200 400 600

0.0

0.1

0.2

0.3

0.4

Peak Model


Per

cent

age


d=119

−600 −400 −200 0 200 400 600

0.04

0.05

0.06

0.07

0.08

0.09

0.10

Peak Model


Per

cent

age


d=50



TFs: MACSHMs: SICER


−600 −400 −200 0 200 400 600

0.0

0.1

0.2

0.3

0.4

Peak Model


Per

cent

age


d=119

−600 −400 −200 0 200 400 600

0.04

0.05

0.06

0.07

0.08

0.09

0.10

Peak Model


Per

cent

age


d=50

WIG file: Normalized Read Density text format⇓

bigWig file: Binary Read Density format



TFs: MACSHMs: SICER


−600 −400 −200 0 200 400 600

0.0

0.1

0.2

0.3

0.4

Peak Model


Per

cent

age


d=119

−600 −400 −200 0 200 400 600

0.04

0.05

0.06

0.07

0.08

0.09

0.10

Peak Model


Per

cent

age


d=50


Primary Analysis ChIPseq Summary

General Guidelines:

Biological Replicates (2x)

INPUT (noIP sample)

20-25 mln reads per sample

Optimize ChIP conditions before library preparation

Minimize experimental variation

Look at the data

’ChIP-seq guidelines and practices of the ENCODE and modENCODE consortia’Landt et al, Genome Res, 2012

’Practical Guidelines for the Comprehensive Analysis of ChIP-seq Data’Bailey et al, PLoS Comp Biol, 2013


Primary Analysis RNAseq

Raw Data

Aligned Data

Read Counts

Condition 1

Read Counts

Condition 2

Differential

Expression



Aligner (Mapper)


...Spliced alignment

Build reference based onknown transcripts

STAR

VERY Fast...(45 million paired readsper hour per processor)

TopHat2

If you want to use CuffLinks...



Raw Data

Aligned Data

Read Counts

Condition 1

Read Counts

Condition 2

htseqCount

Differential

ExpressionDESeqedgeR


Primary Analysis RNAseq Design

Raw Data

Adapted from: Kostadyma M, EBI Workshop


Primary Analysis RNAseq Design

Raw Data

Adapted from: Kostadyma M, EBI WorkshopPapadopoulos GL (IMBB, FORTH) IMBB NGS Seminar 04/12/2015 19 / 28

Interpretation Visualization

GenomeGraphs (R package) Ariadne (GeneViewer)



Local

GenomeGraphs (R package)Ariadne (GeneViewer)



Local

GenomeGraphs (R package)

Online

Ariadne (GeneViewer)



Local

IGV (Integrative Genomics Viewer)


Online




Local



Online

UCSC Genome Browser




Local


Cross Platform (Win, Mac, Linux)Different file formats

Fast


Online

UCSC Genome Browser




Local



Fast


Online

UCSC Genome Browser

Web basedSlow

Comprehensive database




Local



Fast


Online

UCSC Genome Browser

Web basedSlow


Scalechr11:

STS Markers

RatHuman

OrangutanDog

HorseOpossum

ChickenStickleback

RepeatMasker

SNPs (128)

50 kb mm918,780,000 18,790,000 18,800,000 18,810,000 18,820,000 18,830,000 18,840,000 18,850,000 18,860,000 18,870,000 18,880,000 18,890,000 18,900,000 18,910,000 18,920,000

STS Markers on Genetic and Radiation Hybrid Maps

UCSC Genes (RefSeq, GenBank, tRNAs & Comparative Genomics)

G1E DNaseI HS Peaks Rep 2 from ENCODE/PSU

G1E-ER4 Estradiol 24hr DNaseI HS DNase-seq Peaks Rep 2 from ENCODE/PSU

G1E DNaseI HS Signal Rep 2 from ENCODE/PSU

G1E-ER4 Estradiol 24hr DNaseI HS DNase-seq Signal Rep 2 from ENCODE/PSU

CH12 CTCF TFBS ChIP-seq Peaks from ENCODE/Stanford/Yale

CH12 CTCF TFBS ChIP-seq Signal from ENCODE/Stanford/Yale

CH12 p300 (SC-584) TFBS ChIP-seq Peaks from ENCODE/Stanford/Yale

CH12 p300 (SC-584) TFBS ChIP-seq Signal from ENCODE/Stanford/Yale

CH12 Pol2 TFBS ChIP-seq Peaks from ENCODE/Stanford/YaleCH12 Pol2 TFBS ChIP-seq Signal from ENCODE/Stanford/Yale

CH12 c-Jun TFBS ChIP-seq Peaks from ENCODE/Stanford/Yale

CH12 c-Jun TFBS ChIP-seq Signal from ENCODE/Stanford/Yale

MEL CTCF TFBS ChIP-seq Peaks from ENCODE/Stanford/Yale

MEL CTCF TFBS ChIP-seq Signal from ENCODE/Stanford/Yale

MEL GATA-1 TFBS ChIP-seq Peaks from ENCODE/Stanford/Yale

MEL GATA-1 TFBS ChIP-seq Signal from ENCODE/Stanford/Yale

MEL p300 (SC-584) TFBS ChIP-seq Peaks from ENCODE/Stanford/Yale

MEL p300 (SC-584) TFBS ChIP-seq Signal from ENCODE/Stanford/Yale

MEL Pol2 TFBS ChIP-seq Peaks from ENCODE/Stanford/Yale

MEL Pol2 TFBS ChIP-seq Signal from ENCODE/Stanford/Yale

Placental Mammal Basewise Conservation by PhyloP

Multiz Alignments of 30 Vertebrates

Repeating Elements by RepeatMasker

Simple Nucleotide Polymorphisms (dbSNP build 128)

BC050140AK018427

Meis1Meis1

Meis1Gm16140

Meis1AK144295

chr11.2463chr11.2464

chr11.2467 chr11.2474 chr11.2477 chr11.2485chr11.2488

chr11.4792chr11.4794

chr11.4817 chr11.4841chr11.4846

chr11.4858chr11.4862

chr11.4866chr11.4869

chr11.4871

G1E 2

0.15 _

0 _

G1E-ER4 E2 24 2

0.15 _

0 _

CH12 CTCF

CH12 p300 SC-584

CH12 Pol2

CH12 c-Jun

MEL CTCF

MEL GATA-1

MEL p300 SC-584

MEL Pol2

Mammal Cons

2.1 _

-3.3 _

0 -




Local


Online

UCSC Genome Browser

Web basedSlow


Scalechr11:

STS Markers

RatHuman

OrangutanDog

HorseOpossum

ChickenStickleback

RepeatMasker

SNPs (128)

50 kb mm918,780,000 18,790,000 18,800,000 18,810,000 18,820,000 18,830,000 18,840,000 18,850,000 18,860,000 18,870,000 18,880,000 18,890,000 18,900,000 18,910,000 18,920,000

STS Markers on Genetic and Radiation Hybrid Maps

UCSC Genes (RefSeq, GenBank, tRNAs & Comparative Genomics)

G1E DNaseI HS Peaks Rep 2 from ENCODE/PSU

G1E-ER4 Estradiol 24hr DNaseI HS DNase-seq Peaks Rep 2 from ENCODE/PSU

G1E DNaseI HS Signal Rep 2 from ENCODE/PSU

G1E-ER4 Estradiol 24hr DNaseI HS DNase-seq Signal Rep 2 from ENCODE/PSU

CH12 CTCF TFBS ChIP-seq Peaks from ENCODE/Stanford/Yale

CH12 CTCF TFBS ChIP-seq Signal from ENCODE/Stanford/Yale

CH12 p300 (SC-584) TFBS ChIP-seq Peaks from ENCODE/Stanford/Yale

CH12 p300 (SC-584) TFBS ChIP-seq Signal from ENCODE/Stanford/Yale

CH12 Pol2 TFBS ChIP-seq Peaks from ENCODE/Stanford/YaleCH12 Pol2 TFBS ChIP-seq Signal from ENCODE/Stanford/Yale

CH12 c-Jun TFBS ChIP-seq Peaks from ENCODE/Stanford/Yale

CH12 c-Jun TFBS ChIP-seq Signal from ENCODE/Stanford/Yale

MEL CTCF TFBS ChIP-seq Peaks from ENCODE/Stanford/Yale

MEL CTCF TFBS ChIP-seq Signal from ENCODE/Stanford/Yale

MEL GATA-1 TFBS ChIP-seq Peaks from ENCODE/Stanford/Yale

MEL GATA-1 TFBS ChIP-seq Signal from ENCODE/Stanford/Yale

MEL p300 (SC-584) TFBS ChIP-seq Peaks from ENCODE/Stanford/Yale

MEL p300 (SC-584) TFBS ChIP-seq Signal from ENCODE/Stanford/Yale

MEL Pol2 TFBS ChIP-seq Peaks from ENCODE/Stanford/Yale

MEL Pol2 TFBS ChIP-seq Signal from ENCODE/Stanford/Yale

Placental Mammal Basewise Conservation by PhyloP

Multiz Alignments of 30 Vertebrates

Repeating Elements by RepeatMasker

Simple Nucleotide Polymorphisms (dbSNP build 128)

BC050140AK018427

Meis1Meis1

Meis1Gm16140

Meis1AK144295

chr11.2463chr11.2464

chr11.2467 chr11.2474 chr11.2477 chr11.2485chr11.2488

chr11.4792chr11.4794

chr11.4817 chr11.4841chr11.4846

chr11.4858chr11.4862

chr11.4866chr11.4869

chr11.4871

G1E 2

0.15 _

0 _

G1E-ER4 E2 24 2

0.15 _

0 _

CH12 CTCF

CH12 p300 SC-584

CH12 Pol2

CH12 c-Jun

MEL CTCF

MEL GATA-1

MEL p300 SC-584

MEL Pol2

Mammal Cons

2.1 _

-3.3 _

0 -




Local


Online



Interpretation Analysis Software



GALAXY server



GALAXY server

Online collection of widely usedbioinformatics tools

(e.g. FastQC, bowtie.2, CuffLinks,MACS, SICER ...)



GALAXY server



Learn Galaxy...



GALAXY server



Learn Galaxy...

https://galaxyproject.org/




GALAXY server



Learn Galaxy...


Chipster




GALAXY server



Learn Galaxy...


Chipster

User-friendly analysis software forhigh-throughput data (GUI)



Interpretation Downstream Analysis



Genomic Regions Enrichment ofAnnotations Tool (GREAT)




Functional annotation ofcis-regulatory regions





Input: BED file





Input: BED file

Output:List of potential target genes

Peak distribution plotsGeneset enrichment analysis

(GO, Pathways, Phenotypes ...)





Input: BED file




http:

//bejerano.stanford.edu/great/public/html/index.php


http://bejerano.stanford.edu/great/public/html/index.php





Input: BED file




http:


MEME-ChIP







Input: BED file




http:


MEME-ChIP

Perform motif discovery, motifenrichment analysis and clustering







Input: BED file




http:


MEME-ChIP


Input: FASTA file







Input: BED file




http:


MEME-ChIP


Input: FASTA file

Output:Enriched motifs

Motif distribution plots







Input: BED file




http:


MEME-ChIP


Input: FASTA file

Output:Enriched motifs

Motif distribution plots

http://meme-suite.org/tools/meme-chip




http://meme-suite.org/tools/meme-chip

Online Resources ENCODE

Encyclopedia of DNA Elements

Aim: build a comprehensive parts list offunctional elements in the human genome




Aim: build a comprehensive parts list offunctional elements in the human genome

New website is extremely easy to use!


Online Resources GEO

Gene Expression Omnibus




GEO is a public functional genomics data repository





All NGS datasets have to be deposited to GEO prior to publication






All datasets included in GEO are publicly available for academic use







The search function is pretty close to awful








Best way to query the GEO database is by Accession Number(declared in the manuscript, GSE30142, GSM746581)








Best way to query the GEO database is by Accession Number(declared in the manuscript, GSE30142, GSM746581)

FTP access to SRA experiment (Sequence Read Archive)


Online Resources ENA

European Nucleotide Archive




ENA provides a comprehensive record of theworld’s nucleotide sequencing information





It covers raw sequencing data, sequence assembly informationand functional annotation







Large overlap with GEO database


Comprehensive database of genomic features

Systemic understandingof biological processes


Thank you...


introduction to ngs analyses - forth-imbb · introduction to ngs analyses giorgio l papadopoulos...

Documents