Mutation¶
Two main types of genetic studies are population and familial/pedigree studies. In this workflow, individual mutation information is used to determine the relatedness between individuals and data from The Cancer Genome Atlas landmark paper on most common AML mutations is used to reproduce some of the figures in this publication.
![digraph MUTATION {
fontname="sans-serif";
compound="true";
penwidth="0.1";
edge [comment="Wildcard edge",
fontname="sans-serif",
fontsize=10,
colorscheme="blues3",
color=2,
fontcolor=3];
node [fontname="serif",
fontsize=13,
fillcolor="1",
colorscheme="blues4",
color="2",
fontcolor="4",
style="filled"];
subgraph cluster0 {
label="Read data";
edge [comment="Wildcard node added automatic in EG."];
node [comment="Wildcard node added automatic in EG."];
c0_input1 [shape="invhouse",
label="AML genotype"];
c0_input2 [shape="invhouse",
label="AML meta"];
c0_read [label=<Read data<BR/><I>read.delim()</I>>];
c0_check [shape="box",
label=<Validate data<BR/><I>length(), intersect(), order()</I>>];
c0_out [label="list()"];
c0_input1 -> c0_read;
c0_read -> c0_out;
c0_input2 -> c0_read;
c0_out -> c0_check;
{
rank=same;
edge [comment="Wildcard node added automatic in EG."];
node [comment="Wildcard node added automatic in EG."];
c0_input1;
c0_input2;
}
{
rank=same;
edge [comment="Wildcard node added automatic in EG."];
node [comment="Wildcard node added automatic in EG."];
c0_out;
c0_check;
}
}
subgraph cluster1 {
label="reproduce Figure 1a";
edge [comment="Wildcard node added automatic in EG."];
node [comment="Wildcard node added automatic in EG."];
c1_in [shape="invhouse",
label=<aggragate and format data<BR/><I>rbind(), split(), table('tier')</I>>];
c1_meta [shape="invhouse",
label="Meta data \n input$meta"];
c1_mrg [label=<Merge with meta information<BR/><I>merge(x, y)</I>>];
c1_subs [label="subset('tier1')"];
c1_plot [label="plot()"];
c0_out -> c1_in [ltail=cluster0,
lhead=cluster1];
c0_out -> c1_meta [ltail=cluster0,
lhead=cluster1];
c1_in -> c1_mrg [label="x"];
c1_meta -> c1_mrg [label="y"];
c1_mrg -> c1_subs;
c1_subs -> c1_plot;
}
subgraph cluster2 {
label="reproduce Figure 1b";
edge [comment="Wildcard node added automatic in EG."];
node [comment="Wildcard node added automatic in EG."];
c2_in [shape="invhouse",
label=<aggragate and format data<BR/><I>rbind(), split(), table('gene_name')</I>>];
c0_out -> c2_in [ltail=cluster0,
lhead=cluster2];
c2_ordr [label="order(sample_count)"];
c2_subs [label="subset('tier1')"];
c2_head [label="head(100)\ndata.frame(gene_name, tier, sample_count)"];
c2_in -> c2_ordr;
c2_ordr -> c2_subs;
c2_subs -> c2_head;
}
subgraph cluster3 {
label="Load family data";
edge [comment="Wildcard node added automatic in EG."];
node [comment="Wildcard node added automatic in EG."];
c3_input [label="Read genotyping data\n read.delim(genotypes)"];
c3_genot [label=<Summarize genotype freq's <BR/><I>rbind(), table('genotype')</I>>];
c3_compu [shape="box",
label=<create mutation likelihood matrix <BR/><I>matrix(ncol(genotypes), nrow(genotypes))</I>>];
c3_match [shape="box",
label=<compute likelihood scores<BR/><I>number of matching alleles * allele occurence</I>>];
c3_ibdve [shape="box",
label=<Run sliding window over mutations and<BR/>compute 3rd Quantile of mean marker likelihoods<BR/>for different window sizes using<BR/><I>lapply(), summary(), base:::simplify2array(), seq()</I>>];
c3_input -> c3_genot;
c3_genot -> c3_compu;
c3_compu -> c3_match;
c3_match -> c3_ibdve;
}
}](../../_images/graphviz-dac3b6b7939686968eb712b57f39cafbc884c7ef.png)
Diagram of mutation analysis workflow.
Packages and Dependencies¶
There are 2 core packages used in this workflow, which have no dependencies.
Used packages:
- Core: stats, utils
Data¶
The familial data was obtained from the nice people at Genomes Unzipped, who make their own genomic data publicly available.
The dataset we are using comes from the 23andme v2 sequencing service.
Though the individuals are not related, this data can still be used to perform some typical tests carried out on pedigree studies, such as determining “relatedness” between individuals.
| member | dataset id | link |
|---|---|---|
| Daniel MacArthur | DGM001 | http://s3.amazonaws.com/gnz.genotypes/DGM001_genotypes.zip |
| Luke Jostins | LXJ001 | http://s3.amazonaws.com/gnz.genotypes/LXJ001_genotypes.zip |
| Dan Vorhaus | DBV001 | http://s3.amazonaws.com/gnz.genotypes/DBV001_genotypes.zip |
| Caroline Wright | CFW001 | http://s3.amazonaws.com/gnz.genotypes/CFW001_genotypes.zip |
| Kate Morley | KIM001 | http://s3.amazonaws.com/gnz.genotypes/KIM001_genotypes.zip |
| Vincent Plagnol | VXP001 | http://s3.amazonaws.com/gnz.genotypes/VXP001_genotypes.zip |
| Jeff Barrett | JCB001 | http://s3.amazonaws.com/gnz.genotypes/JCB001_genotypes.zip |
| Jan Aerts | JXA001 | http://s3.amazonaws.com/gnz.genotypes/JXA001_genotypes.zip |
| Joe Pickrell | JKP001 | http://s3.amazonaws.com/gnz.genotypes/JKP001_genotypes.zip |
| Don Conrad | DFC001 | http://s3.amazonaws.com/gnz.genotypes/JKP001_genotypes.zip |
| Carl Anderson | CAA001 | http://s3.amazonaws.com/gnz.genotypes/CAA001_genotypes.zip |
| Ilana Fisher | IPF001 | http://s3.amazonaws.com/gnz.genotypes/IPF001_genotypes.zip |
The population study data is from the TCGA consortium publication TCGA, 2013, publication data archive, mutation and annotation (maf), and patient meta data.
License¶
- Copyright (c) 2015 Ieuan Clay based on code from genbench
- Copyright (c) 2015-2016 BeDataDriven B.V. License: GPL version 2 or higher