solarius
R package to interface SOLAR
R package to interface SOLAR
args(solarAssoc)
## function (formula, data, dir, kinship, traits, covlist = "1",
## snpdata, snpformat = "012", cores = 1, ..., verbose = 0)
## NULL
snpformat | snpdata | dat.gen | genocov |
---|---|---|---|
”/” | {A/A, A/T, T/T} | as is snpdata | {0, 1, 2} |
“012” | {0, 1, 2} | {1/1, 1/2, 2/2} | {0, 1, 2} |
“0.1” | [0; 2] | skipped | as is in snpdata |
SOLAR
commands and filesAssociation analysis provided by solarius
package via solarAssoc
function
is based on mga
command of SOLAR
.
The following code shows an exact rules (selected in solarius
)
to conduct an association analysis on 2 cores in parallel.
> mod <- solarAssoc(trait ~ 1, phenodata, snpdata = genodata, kinship = kin, cores = 2)
> mod$solar$assoc$cmd
[[1]]
[1] "load pheno dat.phe"
[2] "load model trait/null0.mod"
[3] "outdir assoc1"
[4] "mga -files snp.genocov -snplists snp.geno-list1 -out assoc.out1"
[[2]]
[1] "load pheno dat.phe"
[2] "load model trait/null0.mod"
[3] "outdir assoc2"
[4] "mga -files snp.genocov -snplists snp.geno-list2 -out assoc.out2"
A list of specific SOLAR
files are used in the calculations,
as described in the following table.
SOLAR file |
solarAssoc argument |
Comments |
---|---|---|
genocov.files | Yes (alternative to snpdata ) |
Data files with genotype-covariates (e.g. file per chromosome) |
genolist.file | No | Meta file with SNP names (internal file derived from genocov.files) |
snplist.files | Yes (alternative to snplist ) |
Meta files with SNP names to analize |
out.dirs | No | Output directories for association models like mga.mod (directory per core) |
out.files | No | Output files to store association results (file per core) |
Two files genocov.files
and genolist.file
are output files
of a group of commands, that convert genotypes to covariates.
These three commands are snp load dat.snp
, snp covar -nohaplos
, snp unload
.
See snpdata2solar
function for more details.
Other three files snplist.file
, out.dirs
and out.files
are custom files by solarius
package
to control association analysis (list of snps to be analyzed, parallel computation).
SOLAR file |
Default value | Parallel computation on 2 cores |
---|---|---|
genocov.files | “snp.genocov” | |
genolist.file | “snp.geno-list” | |
snplist.file | “snp.geno-list” | “snp.geno-list1”, “snp.geno-list2” |
out.dirs | “assoc” | “assoc1”, “assoc2” |
out.files | “assoc.out” | “assoc.out1”, “assoc.out2” |
data(dat50)
mod <- solarAssoc(trait ~ 1, phenodata, snpdata = genodata, kinship = kin)
head(mod$snpf)
## SNP NAv chi pSNP bSNP bSNPse Varexp
## 1 s1 66 0.166224 0.683490 0.144854 0 0.001258
## 2 s2 66 0.851450 0.356142 -0.665254 0 0.080241
## 3 s3 66 0.839794 0.359456 -0.330617 0 0.000000
## 4 s4 66 0.182198 0.669491 0.073773 0 0.000000
## 5 s5 66 0.049818 0.823380 0.344516 0 0.007852
## 6 s6 66 2.882008 0.089574 -0.249210 0 0.051719