Genotype encoded as 0 and 1

[Harisa Muratovic]

Dear Mr. Broman,

The goal of my analysis is to identify epistatic effects between the SNPs I have. My genotype is encoded as zeros and ones, corresponding to recessive homozygotes and dominant ones, respectively. I was able to process the input data as described in the documentation, and am using the following arguments to convert the files into the suitable format:

a=read.cross(format = "csvs", "." ,"./ath_csvs_gen.csv", "./alanine_phe.csv",na.strings="-", genotypes=c("1","0"))

Does the genotype encoding of my data satisfy the requirements, even though the genotypes are already in a numeric format? I am asking since my line of code does not provide any errors, but is taking a long time to output any information (i.e.I havent obtained any output yet). Thinking it could be due to an incorrect format of the geno data or that the read.cross function simply is not made for the genotype encodings I am using.

I thank you for your time.

[Karl Broman]

The encoding of the genotypes shouldn’t pose a problem. I’m not sure whether the slow loading of the data is just due to the size of the dataset, or whether there is some other problem.

karl

On Feb 14, 2024, at 7:47 AM, Harisa Muratovic wrote:



--
You received this message because you are subscribed to the Google Groups "R/qtl discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rqtl-disc+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/rqtl-disc/d5abe019-80c2-40a2-a24a-3ab8c1e3cb99n%40googlegroups.com.

[Harisa Muratovic]

Thank you for the quick response.

Doesnt the inclusion of ids as phenotype impose problems since ids are not phenotype?

"The first column in the genotype data must specify individuals' identifiers, and there must be a column in the phenotype data with precisely the same information (and with the same name). These IDs will be included in the data as a phenotype."

From the documentation source https://search.r-project.org/CRAN/refmans/qtl/html/read.cross.html

Thank you.

[Harisa Muratovic]

I have another question as well.

In which way is the processing of differing cross types changing the output of package's functions? What is the underlying difference for processing a haploid vs a backcross cross type? How would the output and processing change if one were to use genotype data of natural accessions/samples, with no experimental crossing?

[Karl Broman]

Cross type affects the imputation of missing genotype information, at markers and in-between markers, for example in calc.genoprob() and sim.geno().

A backcross and haploids are actually treated the same; they just differ in labels placed on genotypes.

If the data do not come from an experimental cross, then R/qtl is not the appropriate software for the analysis.

karl