R code for analysis of family genome-wide DNA data (e.g. AncestryDNA and 23andme raw DNA data).
Visit the R homepage to learn more about the R statistical programming language. R is a popular system for doing professional DNA data analysis with Bioconductor packages.
This R package, genomology, is for doing genetic geneology and is a bit "hobby-ish" compared to what bioinformatics professionals would typically do with R and Bioconductor.
An example usage of this packge could be determining what percentage of genes a grand-child has received from grand-parents. This can be done with raw DNA data from AncestryDNA or 23andme for a grandparent, a "middle" parent and a grandchild.
You can get package and function documentation from within R by doing:
library(genomology)
help(package="genomology")
In this example:
- three family members DNA data is read (grandma, dad, girl) from different DNA services
- DNA data is merged together with extra information like genes and basepair ranges
- genotypes for lactase persistance (lactose tolerance) regulating gene MCM6 is printed
- estimation of which chromosome regions of girl are descended from grandma
- bar plot by chromosome of how much of girls DNA is inherited from dad's mom (grandma) vs dad's dad
library(genomology)
set.genomology.loci(read.csv("http://ref.castedo.com/dna/loci.csv", comment.char="#"))
grandma <- read.ancestrydna.web("http://example.com/dna-data-grandma.zip")
dad <- read.23andme.web("http://example.com/dna-data-dad.zip")
girl <- read.ancestrydna.web("http://example.com/dna-data-kid.zip")
dna <- grandma[, 1:3]
dna$zone <- chromozone(dna$chromosome, dna$position)
dna$gene <- locus.at(dna$chromosome, dna$position)
dna$gma <- grandma$genotype
dna$dad <- dad$genotype[match(dna$rsid, dad$rsid)]
dna$girl <- girl$genotype
dna <- dna[chromorder23(dna$chromosome, dna$position),]
dna$range <- chromorange(dna$chromosome, dna$position)
subset(dna, gene == 'LCT')
attach(dna)
kidNuc <- allele.assigned(assign.parent(dad, girl), girl)
dadMa <- assign.parent(gma, dad)
kidGrandSign <- assign.grandparents(dadMa, assign.passage(dad, kidNuc))
kidGrandSign[chromosome == 23] <- 1 # dad passes grandma's X to girl
gma.prob <- assign.descent(chromordinate(chromosome, position), kidGrandSign)
tab <- tapply(range/sum(range), list(call=prob.triage(gma.prob), zone=zone), sum)
tab[is.na(tab)] <- 0
barplot(tab, horiz=TRUE, las=1, ylab="Chromosome")
The core mathematical functions used in this package are idescent.prob and recombination.prob.
Documentation is in the package help, but only the PDF (LaTeX) versions have
the mathematical derivations. For convenience, the PDF versions are also online:
Clone this repository and from inside do
R CMD INSTALL package
Please email castedo@castedo.com if you install this package or want a built R package online (which would be easier to install).