zhangh12 / artp2 Goto Github PK

Dear ARTP2 maintainers,

Thanks for ARTP2's read.bed function, I am very happy it is available!

Unexpectedly, for a simple .bed file (created by PLINK v1.9 with --make_bed), read.bed fails, where the genio R package does read it successfully. Below is a reprex with the error message.

I predicted this can be fixed easily, as the solution is laid bare by the reprex :-)

Thanks and cheers, Richel

# Get the data to work on
remotes::install_github("richelbilderbeek/plinkr")
bed_filename <- get_plinkr_filename("demo_assoc.bed")
bim_filename <- get_plinkr_filename("demo_assoc.bim")
fam_filename <- get_plinkr_filename("demo_assoc.fam")

# genio can read the .bed file
genio::read_bed(
  file = bed_filename,
  names_loci = genio::read_bim(bim_filename)$id,
  names_ind =  genio::read_fam(fam_filename)$fam
)

# ARTP2 cannot read the .bed file yet
ARTP2::read.bed(
  bed = get_plinkr_filename("demo_assoc.bed"),
  bim = get_plinkr_filename("demo_assoc.bim"),
  fam = get_plinkr_filename("demo_assoc.fam")
)

The error given is:

Error in `.rowNamesDF<-`(x, value = value) : 
  duplicate 'row.names' are not allowed
In addition: Warning messages:
1: In ARTP2::read.bed(bed = get_plinkr_filename("demo_assoc.bed"),  :
  Duplicated subjects exist in fam file: 
/home/richel/GitHubs/plinkr/inst/extdata/demo_assoc.fam
2: non-unique value when setting 'row.names': ‘1’ 

This gives us the reason for the problem: the .fam file has duplicate IDs. This is reasonable: all individuals should be unique. Well done!

But maybe the function is a bit too overeager. One hint of this may already be that PLINK can work with the .fam file without problems. So, let's take a look at the .fam file:

genio::read_fam(fam_filename)

# A tibble: 3 x 6
  fam   id    pat   mat     sex pheno
  <chr> <chr> <chr> <chr> <int> <dbl>
1 1     1     0     0         1     1
2 2     1     0     0         1     1
3 3     1     0     0         1     1

Aha, all individuals are unique, as all combinations of family ID and within-family ID are unique.

I guess that makes fixing this bug easy :-)

Thanks for this package.

I would like to use this to read plink data.

https://rdrr.io/cran/ARTP2/man/read.bed.html

If I use following code however, I was wondering when reading plink file are the data loaded completely in memory or in chunks or index?

read.bed(bed, bim, fam, sel.snps = NULL, sel.subs = NULL, encode012 = TRUE)

I have plink data for about ~200K individuals.
Thank you.

I got this when I trying to install ARTP2 using package manager in RStudio. What's the compatible R version for ARTP2? Thanks

I learned about GWAS and found a path-based test. From the manual, it seems that it can be executed only in cases and controls study. Is it applicable to continuous traits?

Thank you