\documentclass{standalone}

\usepackage{adjustbox} 

\usepackage[none]{hyphenat} 

\usepackage{tikz} 

\usetikzlibrary{calc,trees,positioning,arrows,chains,shapes.geometric, decorations.pathreplacing,decorations.pathmorphing,shapes, matrix,shapes.symbols}

\tikzset{

>=stealth', punktchain/.style = {

    rectangle, rounded corners, 

    fill = red!06, draw = black,

    text width = 12em, minimum height = 1em, 

    text centered, on chain},

  every join/.style = {->},

  decoration = {brace},

  tuborg/.style = {decorate},

  tubnode/.style = {midway, left = 3pt, text width = 2cm},}

\begin{document} \trimbox{-.5cm -.5cm -.5cm -.5cm}{ 

\begin{tikzpicture}

[node distance = .5cm, start chain = going below,]

\node[punktchain, join] (seqdat) {Download via ftp};

\node[punktchain, join] (qcseq) {Quality control on fastq data \textit{FastQC}, \textit{R}};

\begin{scope}[start branch = venstre, every join/.style = {<-, shorten <= 0.75pt}]

\node[punktchain, on chain = going right, join = by {->}, text width=6em] (fasup) {Upload to NCBI SRA (1.0TB)}; \end{scope}

\begin{scope}[start branch = venstre, every join/.style = {<-, shorten <= 0.75pt}]

\node[punktchain, on chain = going left, join = by {->}, text width=6em] (fasdis) {\small{Discard poor quality samples}}; \end{scope}

\node[punktchain, join] (cleanseq) {Remove adaptor \& low quality sequence \textit{Ea-Utils}};

\node[punktchain, join] (map) {Map reads to reference genome \textit{BWA},\textit{Stampy}};

\node[punktchain, join, fill = green!06] (readgrp) {\raggedright{Add read-group info \\ Remove duplicate reads \\ Build index file \\}\textit{Picard}};

\node[punktchain, join, fill = green!06] (remap2) {Re-map around indels\\ \textit{GATK}};

\node[punktchain, join, fill = green!06] (qcbam) {Bam file QC\\ \textit{GATK}, \textit{Picard}, \textit{R}};

\begin{scope}[start branch = venstre, every join/.style = {<-, shorten <= 0.75pt}]

\node[punktchain, on chain = going right, join = by {->}, text width = 6em, fill = green!06] (bamup) {Upload to NCBI SRA (0.8TB)}; \end{scope}

\begin{scope}[start branch = venstre, every join/.style = {<-, shorten <= 0.75pt}]

\node[punktchain, on chain = going left, join = by {->}, text width=6em, fill = green!06] (bamdis) {\small{Discard poor quality samples}}; \end{scope}

\node[punktchain, join, fill = green!06] (geno) {Variant discovery \& genotyping};

\node [punktchain, below left of = geno, node distance = 2cm, xshift = -1cm, yshift = -0.7cm, fill = blue!06] (hcall) {SNPs and indels:\\\textit{GATK Haplotype Caller}\\\raggedright{Genotype each bam\\Combine gvcfs\\Genotype combined gvcf\\}};

\node [punktchain, below right of = geno, node distance = 2cm, xshift = 1cm, yshift = -0.7cm, fill=blue!06] (gstrip) {Structural variants:\\ \textit{Genomestrip CNV pipeline} \raggedright{Pre-processing\\Discovery\\Genotyping\\}};

[node distance = 0.5cm, start chain = going below,]

\node [punktchain, below left of = gstrip, node distance = 2cm, xshift = -1cm, yshift = -.9cm, fill = blue!06] (genoqc) {Separate QC, annotation \& formatting \textit{GATK}, \textit{vcfTools}, \textit{SNPeff}, \textit{R}};

\begin{scope}[start branch = venstre, every join/.style = {<-, shorten <= 0.75pt}]

\node[punktchain, on chain = going right, text width = 6em, fill = blue!05](genoup) {Upload to NCBI dbVar/dbSNP (13GB)};\end{scope}

\begin{scope}[start branch = venstre, every join/.style = {<-, shorten <= 0.75pt}]

\node[punktchain, on chain = going left, text width = 6em, fill = blue!05](genodis) {\small{Discard poor quality variants and samples}};\end{scope}

% \node[punktchain, fill = blue!06] (combine) {Combine genotype data from \textit{HaplotypeCaller} \& \textit{Genomestrip}};

% \begin{scope}[start branch = venstre, every join/.style = {<-, shorten <= 0.75pt}]

% \node[punktchain, on chain = going right, join = by {->}, text width = 6em, fill = blue!06](comup) {Upload to Zenodo, Fly-var};\end{scope}

\draw[->] (geno.south) |-+(0.5,-1em)-| (hcall.north);

\draw[->] (geno.south) |-+(0.5,-1em)-| (gstrip.north);

\draw[->] (gstrip.south) |-+(0,-1em)-| ([xshift=2pt]genoqc.north);

\draw[->] (hcall.south) |-+(0.5,-1em)-| ([xshift=-2pt]genoqc.north);

\draw[->] ([yshift=1pt]genoqc.west) -- (genodis.east);

\draw[->] ([yshift=-1pt]genoqc.west) -- (genodis.east);

\draw[->] ([yshift=1pt]genoqc.east) -- (genoup.west);

\draw[->] ([yshift=-1pt]genoqc.east) -- (genoup.west);

% \draw[->] ([xshift=-2pt]genoqc.south) -- (combine.north);

% \draw[->] ([xshift=2pt]genoqc.south) -- (combine.north);

 \draw[tuborg, decoration={brace}] let \p1=(seqdat.north), \p2=(map.south) in ($(-5.5, \y2)$) -- ($(-5.5, \y1)$) node[tubnode] {\centering{Sequencing data \textit{*.fastq}}};

\draw[tuborg, decoration={brace}] let \p1=(readgrp.north), \p2=(geno.south) in ($(-5.5, \y2)$) -- ($(-5.5, \y1)$) node[tubnode] {Sequence alignment data \textit{*.bam}};

\draw[tuborg, decoration={brace}] let \p1=(gstrip.north), \p2=(genoqc.south) in($(-5.5, \y2)$) -- ($(-5.5, \y1)$) node[tubnode] {Genotype data \textit{*.vcf}};

\end{tikzpicture}}

\end{document}

%%% Local Variables: 

%%% mode: latex

%%% TeX-master: t

%%% End: