|
| 1 | +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 2 | +\section{Modularization} |
| 3 | +{ |
| 4 | + \usebackgroundtemplate{ |
| 5 | + \vbox to \paperheight{\vfil\hbox to \paperwidth{\hfil\includegraphics[height=.7\paperheight]{misc/bricks.png}\hfil}\vfil} |
| 6 | + } |
| 7 | + \frame{ |
| 8 | + \frametitle{Modulerizing Snakefiles} |
| 9 | + \begin{mdframed}[tikzsetting={draw=white,fill=white,fill opacity=0.8, |
| 10 | + line width=0pt},backgroundcolor=none,leftmargin=0, |
| 11 | + rightmargin=150,innertopmargin=4pt,roundcorner=10pt] |
| 12 | + \tableofcontents[currentsection,sections={1-4},hideothersubsections] |
| 13 | + \end{mdframed} |
| 14 | + } |
| 15 | +} |
| 16 | + |
| 17 | + |
| 18 | +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 19 | +\begin{frame} |
| 20 | + \frametitle{What is this about?} |
| 21 | + \begin{question}[Questions] |
| 22 | + \begin{itemize} |
| 23 | + \item Are monolithic Snakefiles feasible? |
| 24 | + \item How can we put a Snakefile in pieces? |
| 25 | + \end{itemize} |
| 26 | + \end{question} |
| 27 | + \begin{docs}[Objectives] |
| 28 | + \begin{enumerate} |
| 29 | + \item Introduce maintainability concepts |
| 30 | + \item Introducing modularization strategies |
| 31 | + \end{enumerate} |
| 32 | + \end{docs} |
| 33 | +\end{frame} |
| 34 | + |
| 35 | +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 36 | +\begin{frame} |
| 37 | + \frametitle{Intro to Maintainability} |
| 38 | + \begin{question}[Take a guess:] |
| 39 | + \begin{itemize} |
| 40 | + \item How big can a workflow be (lines of code)? \pause A few hundred to thousand lines. |
| 41 | + \item Does it then make sense to keep it all in one file? |
| 42 | + \end{itemize} |
| 43 | + \end{question} |
| 44 | + \pause |
| 45 | + \begin{hint} |
| 46 | + The ideas behind code modularization are:\newline |
| 47 | + Increase readability, \pause enable parallel devolpment \pause ease testing and debugging. |
| 48 | + \end{hint} |
| 49 | +\end{frame} |
| 50 | + |
| 51 | +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 52 | +\begin{frame}[fragile] |
| 53 | + \frametitle{Dividing Snakefiles} |
| 54 | + Our Snakefile hardly merits modularization, but we could move all mapping-realted stuff to an extra file, e.\,g.: |
| 55 | + \begin{lstlisting}[language=Python,style=Python] |
| 56 | +include: "rules/mapping.smk" # will include this Snakefile |
| 57 | + \end{lstlisting} |
| 58 | + \pause |
| 59 | + \begin{docs} |
| 60 | + \begin{itemize}[<+->] |
| 61 | + \item we place our modules into \altverb{rules} |
| 62 | + \item the suffix \altverb{.smk} is convention, too. |
| 63 | + \end{itemize} |
| 64 | + \end{docs} |
| 65 | +\end{frame} |
| 66 | + |
| 67 | +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 68 | +\begin{frame}[fragile] |
| 69 | + \frametitle{How a good Snakefile \emph{should} look like} |
| 70 | + \begin{lstlisting}[language=Python,style=Python] |
| 71 | +from snakemake.utils import min_version |
| 72 | + |
| 73 | +# require a minimum version |
| 74 | +min_version("8.10.7") |
| 75 | + |
| 76 | +# may point to samples, contains sciency parts |
| 77 | +configfile: "config/config.yml" |
| 78 | +# entry point for the workflow documentation |
| 79 | +report: "report/workflow.rst" |
| 80 | + |
| 81 | +include: "rules/commons.smk" |
| 82 | +include: "rules/qc.smk" |
| 83 | +include: "rules/utils.smk" |
| 84 | +include: "rules/ref.smk" |
| 85 | +include: "rules/alignment.smk" |
| 86 | +include: "rules/diffexp.smk" |
| 87 | + |
| 88 | +inputdir = config["inputdir"] |
| 89 | + |
| 90 | +rule all: |
| 91 | + input: |
| 92 | + # generates targets |
| 93 | + rule_all_input(), |
| 94 | + \end{lstlisting} |
| 95 | +\end{frame} |
| 96 | + |
| 97 | +\begin{frame}[fragile] |
| 98 | + \frametitle{\HandsOn{Practical Modularization}} |
| 99 | + Try creating this major Snakefile: |
| 100 | + \begin{lstlisting}[language=Python,style=Python] |
| 101 | +from snakemake.utils import min_version |
| 102 | + |
| 103 | +# which is the current snakemake version? |
| 104 | +min_version(<the current snakemake version>) |
| 105 | + |
| 106 | +rule all: |
| 107 | + input: |
| 108 | + #our targets |
| 109 | + |
| 110 | +include: "rules/mapping.smk" |
| 111 | +include: "rules/index.smk" |
| 112 | +include: "rules/call.smk" |
| 113 | +include: "rules/plots.smk" |
| 114 | + \end{lstlisting} |
| 115 | +\end{frame} |
| 116 | + |
| 117 | +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 118 | +\begin{frame}[fragile] |
| 119 | + \frametitle{Alien Modules} |
| 120 | + \Snakemake allows to "borrow" from CWL or Nextflow, too: |
| 121 | + \begin{lstlisting}[language=Python,style=Python] |
| 122 | +rule samtools_sort: |
| 123 | + input: |
| 124 | + input="mapped/{sample}.bam" |
| 125 | + output: |
| 126 | + output_name="mapped/{sample}.sorted.bam" |
| 127 | + params: |
| 128 | + threads=lambda wildcards, threads: threads, |
| 129 | + memory="4G" |
| 130 | + threads: 8 |
| 131 | + cwl: |
| 132 | + "https://github.com/common-workflow-language/workflows/blob/" |
| 133 | + "fb406c95/tools/samtools-sort.cwl" |
| 134 | + \end{lstlisting} |
| 135 | +\end{frame} |
0 commit comments