Hello everyone,

I have: CHR START END EffectAllele RefAllele columns (hg38)

I want: CHR START END EffectAllele RefAllele RS_ID

I would like to get the SNP rs id using chromosome coordinates, please let me know how I can do this?

Thanks in Advance

Here's an example I originally posted to BioStars. If you already have the allele information you can drop that from the attributes.

library(biomaRt)
## Use the default ENSEMBL Variation Mart & Human dataset
snpMart = useEnsembl(biomart = "snps", 
                 dataset = "hsapiens_snp")

## Create an example set of coordinates as a dataframe
SNP_M <- data.frame(CHR = c(1,1), START = c(10020, 10039), END = c(10020, 10039))

## Combine these into the format chr:start:end
## It's important to include the end even if it's a single base, 
## otherwise it searches to the end of the chromosome
coords <- apply(SNP_M, 1, paste, collapse = ":")
coords
#> [1] "1:10020:10020" "1:10039:10039"

## Submit the query
getBM(attributes = c('refsnp_id', 'chr_name', 'chrom_start', 'chrom_end', 'allele'),
      filters = c('chromosomal_region'), 
      values = coords, 
      mart = snpMart)  
#>     refsnp_id chr_name chrom_start chrom_end allele
#> 1 rs775809821        1       10020     10021   AA/A
#> 2 rs978760828        1       10039     10039    A/C

Hi,

Here is an alternative that uses a SNPlocs package:

library(SNPlocs.Hsapiens.dbSNP151.GRCh38)

my_ranges <- GRanges(c("1:1-10050", "2:1-10050"))

## The first call to snpsByOverlaps() takes a while. Subsequent calls are fast.
snpsByOverlaps(SNPlocs.Hsapiens.dbSNP151.GRCh38, my_ranges)
# UnstitchedGPos object with 3 positions and 2 metadata columns:
#       seqnames       pos strand |    RefSNP_id alleles_as_ambig
#          <Rle> <integer>  <Rle> |  <character>      <character>
#   [1]        1     10039      * |  rs978760828                M
#   [2]        1     10043      * | rs1008829651                W
#   [3]        2     10026      * | rs1366167113                R
#   -------
#   seqinfo: 25 sequences (1 circular) from GRCh38.p7 genome

Difference with the result returned by biomaRt probably due to the latter being based on a different dbSNP build.

To get the ref/alt alleles (works in BioC devel only):

library(BSgenome.Hsapiens.UCSC.hg38)
genome <- BSgenome.Hsapiens.UCSC.hg38
seqlevelsStyle(genome) <- "NCBI"

snpsByOverlaps(SNPlocs.Hsapiens.dbSNP151.GRCh38, my_ranges, genome=genome)
# UnstitchedGPos object with 3 positions and 5 metadata columns:
#       seqnames       pos strand |    RefSNP_id alleles_as_ambig genome_compat
#          <Rle> <integer>  <Rle> |  <character>      <character>     <logical>
#   [1]        1     10039      * |  rs978760828                M          TRUE
#   [2]        1     10043      * | rs1008829651                W          TRUE
#   [3]        2     10026      * | rs1366167113                R          TRUE
#        ref_allele     alt_alleles
#       <character> <CharacterList>
#   [1]           A               C
#   [2]           T               A
#   [3]           A               G
#   -------
#   seqinfo: 25 sequences (1 circular) from GRCh38.p7 genome

See ?snpsByOverlaps in the BSgenome package for more information.

Best,

H.