Nuts and Bolts of Next Generaon Sequencing Analysis at ACCRE · 2017-11-09 · Nuts and Bolts of Next Generaon Sequencing Analysis at ACCRE Thomas Stricker MD/PhD Assistant Professor,

NutsandBoltsofNextGenera1onSequencingAnalysisatACCRE

ThomasStrickerMD/PhDAssistantProfessor,PMI

11/09/2017

NextGenera1onSequencingIllumina Seqeuncing Technology

DNA–thegene+ccode• DNAisadoublestrandedpolymerof4bases(A,T,C,G)

• Theorder(sequence)ofA,T,C,Gisthegene1ccode

• AalwayspairswithTontheoppositestrand,andCalwayspairswithG

• EnzymescalledpolymerasesmakecopiesofDNAbytakingasinglestrandofDNA,andthenaddingA,T,C,Gaccordingtothebase-pairingrules

Sanger(modbyLeeHood)• Sequencingbysynthesis

• MixmanycopiesofthesameDNAmolecule,polymerase,ATCGs,andasmallamountofflourescentlylabeledATCGthatareterminated

• Terminatedbasesstopextension

• Separatebasedonsize

1.  In vitro amplification, ‘cloning’

2.  Flow cell based sequencing by synthesis

3.  A draft of the human genome

Illumina Seqeuncing Technology

WhatHappened?

TermsandDefini1ons

•  Singleorpairedend•  Lane•  Barcodeorindexes

•  Library–Thebaseunitofprepara1on.PoolofDNAmoleculesthatareseq.

•  Sample–Asingleindividual

TermsandDefini1ons•  FASTA=sequencefile•  Fastq=sequence+qualityfile•  BAM=alignmentfileinbinary•  VCF=variantcallfile•  Index=allowsrapidlookupoffasta,bam,vcf•  MAF=muta1onannota1onfile•  GTF/GFF=informa1onaboutgenestructure•  Intervallist=likeBED,uniquePicard(usepicardBedteoIntercalto

generate)•  BED=

Contents of the human genome [CC-BY-SA-3.0] Steve Cook

SequencingandGenomeSize

S. cerevisiae = 12 million bp C. elegans = 100 million bp Drosophila m. = 130 million bp D. rerio = 1.4 billion bp M. musculus = 2.1 billion bp H. sapiens = 3 billion bp

Genomic DNA

Pros: Large capture space Copy Number and Translocations Variants outside baits might be sampled Cons: Larger amount of DNA necessary (>250ng) Sequencing inefficient Library builds slow, add to turnaround time

TargetEnrichment–HybridCaptureApproach

ALK APC BRAF BRCA1 BRCA2

ERBB2 EGFR FLT3 HRAS IDH1 IDH2 JAK2 KIT KRAS MET MPL MTOR MYC

NF1 NRAS PDGFRA PIK3CA PTEN PTPN11 TP53

Pros: Quick, efficient, low sample amounts (10ng) Cons: Number of genes limited Uneven PCR = uneven coverage No copy number or translocations Variants outside amplicon not sampled

TargetEnrichment–AmpliconApproach

RNAseq

•  Splice aware aligner build splice junctions on the fly

•  Isoform specific quantification •  Ab initio gene/isoform discovery •  Fusion genes •  Genotype (SNVs only)

Gene A GeneB Exon A Exon B Exon A Exon C

Exon B

(ChIP-seq)• Methodofiden1fyinggenome-widebindingsitesforapar1cularTF,bypurifyingTF-boundDNAandsequencingittolocategenomicTFbindingregions• Canbeusedonbothendogenousandexogenousproteins• Endogenousdetec1ondependsonreliablean1bodyforIP,althoughtaggedproteinscanbeusedforthesamepurpose

Contents of the human genome [CC-BY-SA-3.0] Steve Cook

SequencingandGenomeSize

S. cerevisiae = 12 million bp C. elegans = 100 million bp Drosophila m. = 130 million bp D. rerio = 1.4 billion bp M. musculus = 2.1 billion bp H. sapiens = 3 billion bp

GenomeandGeneVersions

•  Hg19/GRCh37orGRCh38•  Refseq,Genbank,Ensembl…

FASTQfilesandQC

Pre-processingSteps

•  1)QCrawdatawithfastqc•  2)Trimtoremoveadaptors/lowqualitysequencing

•  3)QCclippeddatawithfastqc•  4)Summarizeoutputwithmul1qc

FASTQandQC

•  FASTQC•  /home/strickt2/scripts/qc/fastqc.pl

•  GotoFASTQCandMul1QCsummaries

RNAseqExpressionAnalysis

•  Splice-awarealignment–bamfile–  Tophat2,STAR,RSEM

•  PicardandRSeQCalignmentQC•  SummarizeQC•  Assignreadstotranscripts,producecountfiles

– HtSeq,featurecounts,cufflinks•  Collectintomatrix•  Differen1alExpressionwithDESeq2,limma/voom,EdgeR,others,mostlyinR

Splice-awarealignment

•  STAR--2step–  /home/strickt2/scripts/rnaseq/star.step1.pl

•  BamFile–  /home/strickt2/SEQanalysis/samtools/bin/samtoolsview-H/data/strickt2/578/clip/578-AG-49.clip/Aligned.sortedByCoord.out.bam

•  SummarizedmetricQC•  Featurecounts

–  /data/strickt2/578/clip/new_count/final.transcript.table.txt

Genotyping--Germline

IGV – Genotyping

AddReadGroups

•  ReadGroups=setofreadsgeneratedfromasinglesequencingrun

•  ID=flowcellID+Lane•  PU=plahormunit=flowcellID+Lane+SampleID

•  PL=Illumina,IonTorrent,etc.•  LB=LibraryID

/home/strickt2/scripts/rnaseq_genotyping/final.stricker.merge.rg.mdup.pl

Genotyping--Germline

hips://sokware.broadins1tute.org/gatk/best-prac1ces/

HaplotypeCaller/JointCaller


VCF–variantcallfile•  /data/strickt2/3102/3102-CLA-67.3102.split.vcf

FilterGenotypes•  Recalibrate(VQSR)orhard-filter

–  Recalibra1onisrecommendedforgermlinehumanvariantsfor30ormoreexomes,ormodelorganismsforwhichtheappropriatetrainingsetsexist

•  Recalibra1onhastoberuntwice–onceforSNPsandonceforpolymorphisms

•  Trainamodelofvariantsta1s1cs(QD(qualityscoreoverdepth),MQ(RMSofmappingquality),SB(strandbias),etc.)forknownvariants

•  Usethatmodeltodeterminetheprobabilitythatothervariantsaretrue.


VQSR–Recalibrate


SNPs

Indels

VQSR--Recalibrate


SNP Indel

HardFilteringhips://sokware.broadins1tute.org/gatk/best-prac1ces/

HardFiltering•  QD=QualByDepth=Variantconfidence/Depth•  FS=FisherStrand=Phred-scaledp-valuefromFETstrandbias•  MQ=RMSMappingQuality•  MQRankSum=MappingQualityRankSumTest=CompareMQforreadswithrefvs.

readswithaltviaMann-Whitney•  ReadPosRankSum=ReadPosRankSumTest=Mann_Whitneyfordistanceofvariant

fromendofread.Variantsonlyseenatendofreadslikelyerrors.•  SOR=StrandOddsRa1o=strandbias.

•  READ:•  hips://sokware.broadins1tute.org/gatk/documenta1on/ar1cle.php?id=3225

•  hips://sokware.broadins1tute.org/gatk/documenta1on/ar1cle.php?id=6925and


Annota1ons

•  Annovar,VEP,others•  Availabledataresources

– Humanvaria1on(dbSNP,ExAC,gnomAD,EVS,etc.)

– Clinicalsignificance– ClinSig,CancerHotspots,others

– Measuresofconserva1onordeleteriousness•  GERP,CADD,SIFT,Polyphen,Provean,etc

•  Annota1ontablesandRadarplots

Somatic Genotyping –Tumor-Normal Contamination

Tumor-Normal Contamination

Framptonet.al.hips://www.nature.com/ar1cles/nbt.2696.

Soma1cMuta1onCalling•  Tumor-NormalPairs

OtherAnalyses

•  ChIPseq/ATAC-seq/etc.•  Methyla1on/Bisulfitesequencing•  FusionIden1fica1on•  Noveltranscript/isoformiden1fica1on•  CopyNumberAltera1ons

– BothGermlineandSoma1c

•  Allele-Specificexpression

FusionIden1fica1onCCL14:PCGF2USP22:PTPRDNR2C2:RFTN1ESR1:AKAP12ESR1:C6orf211PAK1:RP11-807H22.7ESR1:C6orf97THSD4:LRRC49FOXK2:ZBTB40CES7:FANCD2ESR1:C6orf97PBRM1:NKIRAS1PTPRN2:CYP3A5SPOP:SWAP70DNAJA3:PTPN2NR3C2:FSTL5OSBPL2:CDH4ROR1:LRP8SSH1:FOXN4NRIP1:AF127936.7NCOR1:WDR16CCL4L2:CCL4ERLIN1:PTENFAM188A:PIP4K2APBX1:RP11-705O24.1STK32B:CLEC16AMLL3:ANKRD36

A B

Gene A GeneB Exon A Exon B Exon A Exon C

Exon B

TheFuture

•  SingleCellRNAseq– ddSEQ,10XChromium(JeffRathmell),InDrop(LauLab),andSmartSeq2(Mallal)

•  Synthe1cLongReads– haplotypingandcopynumber/structuralvariantanalysis

•  Spark-enabledGATK4•  Costeffec1veexomeandwholegenomesequencing=datadeluge

Nuts and Bolts of Next Generaon Sequencing Analysis at ACCRE · 2017-11-09 · Nuts and Bolts of Next Generaon Sequencing Analysis at ACCRE Thomas Stricker MD/PhD Assistant Professor,

Documents