This bioinformatics lesson is brought to you by the letter 'W'

Today's bioinformatics lesson is brought to you by the letter 'W'

by

Keith Bradnam

Image from flickr.com/91619273@N00/

Today's bloinformaties lessonis brought to you by the letter 1W1

Image from flickr.com/91619273©NO0/

Wis for Workflowsis for Workflows

A typical bioinformatics workflow

Illumina data(FASTQ format)

Remove adapter contamination

A typical bioinformatics workflow

Remove adapter contamination

A typical bioinformatics workflow

Illumina data(FASTQ format)

Remove adapter contamination

scythe cutadapt trimgalore

skewer Btrim

Trimmomatic

A typical bioinformatics workflow

Remove adapter contamination

scythecutadapttrimgaloreskewerBtrim

Trimmomatic

A typical bioinformatics workflow

Illumina data(FASTQ format)

Remove adapter contamination

scythe cutadapt trimgalore

skewer Btrim

Trimmomatic

Lots of tools you could use!

A typical bioinformatics workflow

Lots of toolsyou could use!

Remove adapter contamination

scythecutadapttrimgaloreskewerBtrim

Trimmomatic

Trim reads for low quality bases

sickle Qtrim

FastQC FastX

PRINSEQ Trimmomatic

Trim reads for low quality bases

sickleQtrim

FastQCFastX

PRINSEC)Trimmomatic

Map reads to genome/transcriptome

BWA Bowtie TopHat SHRiMP BFAST MAQ

From ebi.ac.uk/~nf/hts_mappers/

There are a lot of read mappers out there!

From ebi.ac.uk/-nf/hts_mappers/ H I S A T •-JAGuaR • -BWA-PSSM • - -MOSAIK •- - - - - -Hobbes2 •CUSHAW3 a-

NextGenMap •Subread/Subjunc •CRAC •-SRmapper •-GEM •STAR •ERNE •-BatMelh •-BLASR a-YAHA •

SeciAlto •Batmis •There are a lot ofDynMaPp O S A •

ContextMap •-as?n1 •-RUM a_read mappers out there! StampydrFAST •-Bismark •-•-

MapSplice a- REAL a--BS-Seeker a-- - B S - S e e k e r 2 - ••SupersplatliceMapRAT • - B R A T - S W -•-BFAST •-

segemeht •-GNUMAP •-GenomeMapper •-mrFAST • • - mrsFAST m r s FA S T- L i l t r a - -• - - - -PerM • - - - - - - --RNA-Mate • - - - X-Mate a- - - - SBSMAP • - - - - S p l a z e rRazerS • --•- -MicroRazerS - • - - • RazerS3SHRIMP a — —• SHR1MP2 -•BWA s - - • BWA-SWCloudBurst •ProbeMatch • • W H A M - •

TopHat a- T o p H a t 2 -•-Bowlie •- B o w t i e 2 •-MOM 4-PASS •- P A S S - b i s - -•Slider • - - -Slider-II-()PALMA •SOCS "-MAO •SegMap •ZOOM •PalMaN a-RMAP •SOAP • —SOAP2- -•BWT-SW • - - S O A P S p l i c e - -•

Blat a-SSAHA •

GMAP •Exonerate •Mummer 3 •

ELAND • GSNAP- a-

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015Years

Map reads to genome/transcriptome

BWA Bowtie TopHat SHRiMP BFAST MAQ

From ebi.ac.uk/~nf/hts_mappers/From eloi.ac.uki-ntiGnotdrni et Atft.- 2 c 1 4 . 1.5auppl 9:512hitk.,:,www.bicrileckentrakuoiryt41-2105/75.•9•512

HISATJAGuaIR - -Bw •A-PSSM - - - -M0-A1K

ApproachARYANA: Aligning Reads by Vet Another

Milad Gnoliimi • r, Arjean kba:: ' ', Ali Sharifiviv:1-• .44, Harritireza (..hitsaz Merio. . ..ignit5.AbstractPitTsburgh , PA, 1..,'S A 31 March - OS April 20.4

iert)m Ric:COM8-Seq: Fourth Annual RkC(....V/111 Satellite 'Workshop or) Massively Parallel SequencingMotivation: Although there are

'•'--Aarly cihretent aigorithms anc software rook br Nigning sequencing reacio s rgappeo s,Fo./pnce search is far from soiven Strong Interest in fast alignrrien:- is hest 1.1,1pc7e0 in the S V or .7tm foraigorithms ',V- rh beperri on fast a rid accurate alignment.

anclitiort de now? assembty of neat-Generaton iPet. enring lng readequites fast overiap-layriur-concensustie Innoczmve competition on a going a roller:ton of reads to a giver database df reference genomes. In

-f_ultra- • -Contribution: I'le introt-Lre ARvANA. a fast gappec rear! aligner developed on Me biss of iilleA incleA•ingnisastr,_cture with a co-ripletely neoo a ighrrent eng OP that rh.akes it signrfiramly faster than 7hree other aligner's:

Sowtie2, BMA anti SegAirt), w tn comparable Gen -t,c-.: ty ant: acruracy. Instead of thp orne-consurning t-haricraciong:vac:et:ores ''L,!• handhing rntsrnatrtx5,s, ARYANIA come; with thp sese-anO-exten0 aigorIMmir framework ano a5lonificantly IrnPrOved mth

efficiency by Integrong riNpi algorithmic tetirnidt.el incluongdynamAr seer: seteCtion,nin 'ectional spec eltensiort reset-4.rep hash tables ano gap-filling cAnynn•nir brogsarnming. As thp reac length _ - -

increases ARYA-V/A•.!T Itioeflorny in terms of speed ana ahgnment rate becomes more evelent. This is in perfect',lakes At par)/ to deveion mission-specie Nigners for other appiications using ARVANA engine.harmony 4vith the iFeli lit'ngth trena as :he seci4enclnig Technologies evohie Ihe algorithmc plaTform of ARYANA

introduction

Availability: ARYAN.4 compip7e source rexie can he obrairteil from kittp.//gitbubcOrnlar)'ana-aligner

i:vt-ty liv:ns cell carries a hatA4 offnre consisting or several used a laborious hierarchil process to divide the gertorne

thnuNand itl r

billitms of characteni with answers to many into srnalier. coveg tam while the Celera (;i-siolnics firmvital qumlions_ .1-11.mnin efforts to decipher that hook has replaced that b rin

y a trnnputational sequence-assembly soli -

Islernatio,:rat ilnynan Genolne ..eq.ite-ncing Conxort,Lion

gained increasing :rloitivntlint since /953 WhtiL the double ware applied to the data geneated front bhoelly shreddedhelical structure 011)NA was discovered- 'twenty years (shotgun) whole gentorte 17,.ti:. 'the automated Sanger

Liter. W.. Gilbert and A. Maxarn react the nrst 2,1-tit...It-atter r

method was the gold standard fin- about two dettleN, asword of the book [I]. svhen II Sanger and his tsolleasties the.first *-ene.,-ntieor or 021i/A xecitiencing. until iecreasing

application of labeled dideoxynucleotide triphosphatex volome of en-or free genomir information can%ed miler-were dmeloping anothm sequenting method based on the demand for la.,,t and inexpensive methods to produce high

I I

that act ;IS chain terminators in a PC.R rmclior: /2,3...

gence of new technologies. the so tailed Nett-Geno-rainn Idrearn of reading the hunzari honk f e was rtallaed hyAbout three decades after the firn ONA vegurnLing, Sequericisv OVG,S)

.-1, paradigrn shih in both the experimental technititieli 2 0 1 3 2 0 1 4 2 0 1 5completion of the t 3 I li t h e frulnan genrmre profect (4-61, rhe and computational Inettulth octurred

doe to the transition

SSAHA • -II B l o t •-_

Ftli 1st ca' Aut'O' iniblniran 1 avaiklii‘ 41 MI' (–CI a? V* artfig•.

rit:ctir;s1P, eye ive Sanger mate-paired reads t -, -41t7 to

• coeirsgt:,-,1,vi, i,),:kly•ieri?itt,ari,

relmenre gerunnes, such as the human genotr , or more

hvananli J-Ktruto a' V are Sarrt-tunnow tr-eas, tat,

t tore-.4.0,7 f4,,ati,

than 2000 prokitryotex- toilvar), nes and Archaea. lamg,

to the NGS tec:hnologies and also ;Availability of finished2001 2 0 0 0 Wattled Central '''''..•„Nzvoetr - - - - — -ec the crtPrta 44..0 ,,,,t,:.0.,. a.,....„.0,,,elun.:06,z, kx...,0_,-;:t:eC—rnOrdo.Ercfo;CerretnseS:0;xa:13'stect'AL:i.deelat;,,13,17,a5Vt. GISrbtco,„.-"•amoeue? aro% x,,,, (-1'sYl't “:""Mort$ Fttec r,... -0 -?D14 ',1C.4,Tr'l elow:ccrseitv..43P.Ittfrtfct 'NI a 61 Lt)&-. ACUIS ark* a rnkozo imat, re :errra o' rPt .v•nit

el, A

(611;Bloinformatics

Filter for uniquely mapped reads

SAMtools Picard GATK Unix

Filter for uniquely mapped reads

SAMtoolsPicardGATKUnix

Filter for high quality alignments

SAMtools Picard GATK Unix

Filter for high quality alignments

SAMtoolsPicardGATKUnix

Data suitable for final analysis

Data suitable forfinal analysis

Some questions you should ask yourself…Some questions you should ask yourself..

Wis for 'Why?'is for 'Why?

Why are each of these steps needed?Why are each of these steps needed?

Why should I use tool 'X' at this step?Why should I use tool X' at this step?

Wis for 'What?'is for 'What?'

What is the effect on running each step?What is the effect on running each step?

What is a good result?What is a good result?

The effect of applying many 'bioinformatics axes'

Illumina data(FASTQ format)

2 FASTQ files

Files are ~6.5 GB

52.5 million reads total

The effect of applying many1bloinformatics axes'

IIlumina data(FASTQ format)

2 FASIQ files52.5 million reads total

Files are ,-,64.5 GB

Remove adapters & trim

50.1 million reads

Remove adapters & trim

50.1 million reads

Align to transcriptome with Bowtie

35.8 million reads map

Align to transcriptome with Bowtie

35.8 million reads map

Filter for uniquely mapped reads

31.4 million reads align uniquely

Filter for uniquely mapped reads

31.4 million reads align uniquely

Filter for high quality alignments

22.7 million reads have alignment scores of zero

Filter for high quality alignments

22.7 million reads have alignment scores of zero

Data suitable for final analysis

Reduced data from 52.5 to 22.7 million reads

Data suitable forfinal analysis

Reduced data from 52.5 to 22.7 million reads

It can be helpful to know how the different steps in a workflow reduce your data

It can be helpful to know how the differentsteps in a workflow reduce your data

One final tip…One final tip...

ls -ltris l t r

Run this command after every step of a workflowRun this command afterevery step of a workflow

Let's you see whether output files were actually created

Let's you see whether output fileswere actually created

Let's you see whether output files contain any data

Let's you see whether output filescontain any data

Most recently modified files will be at bottom of your terminal windowMost recently modified files will beat bottom of your terminal window

The endThe end