From gene symbol list to their coordinates

Question

[cfarmeri]

Hi, Biostars. I have a gene symbol list (plain text file) as following example.

=============

Nanog

Sox2

Dnmt3l

Rex2

・

・

・

=============

From this, I would like to get these gene's coordinate as following.

=============

Nanog,Chr6:122707489-122714633,+

Sox2,Chr3:34650405-34652461,+

Dnmt3l,Chr10:78041947-78063622,+

Rex2,Chr4:147021850-147060794,+

・

・

・

=============

what the most simple way of you know solves above ?

Could you please suggest me anything and everything!!

Answer 1

[venu]

One way is using biomaRt package from Bioconductor. If you don't want to use the this method, the other way is

• Download appropriate reference genome GTF file from Ensembl
• Convert your Gene symbols to Ensembl ids (Id Conversion online)
• Extract Gene coordinates from GTF file for your genes.

Both methods require little bit of programming knowledge (R knowledge for first method and some simple unix commands for second). I think online tool Ensembl Biomart provide this information, though I've never used it.

[Emily]

The BioMart online tool is probably the easiest way to do it. There's a video tutorial to get you started. Just filter by your list of gene names (ID list: Associated gene name) and get the coordinates as attributes.

[rbagnall]

Thanks Emily, that is indeed very simple. Do you know how the gene start and end position is defined? Transcription start/end, or perhaps coding start/end?

[Emily]

5' transcription start of the most 5' transcript and 3' transcription end of the most 3' transcript

[cfarmeri]

Thanks venu. I solved my problem by biomaRt package!! So thanks !!! The other ways you suggest also look very useful.

Answer 2

[rbagnall]

Using the UCSC table browser Select the following options:

genome: mouse
assembly: Dec. 2011 (GRCm38/mm10)
group: Genes and Gene predictions
track: RefSeq Genes
table: refGene
region: genome
Identifiers (names/accessions): paste list {then paste the list of gene names}
output format: selected fields from primary and related tables
output file: results.txt

Then click 'get output' and in the following window select:

name
chrom
strand
txStart
txEnd
name2

This gives the following output in results.txt:

chrom strand txStart txEnd name2
chr3 + 34649994 34652460 Sox2
chr4 + 147021849 147060799 Rex2
chr6 + 122707564 122714633 Nanog
chr6 + 122707564 122714633 Nanog
chr6 + 122707488 122714633 Nanog
chr6 + 122707564 122714633 Nanog
chr10 + 78042286 78063622 Dnmt3l
chr10 + 78049958 78063622 Dnmt3l
chr10 + 78055334 78063622 Dnmt3l
chr10 + 78055334 78063622 Dnmt3l
chr10 + 78055334 78063622 Dnmt3l
chr10 + 78049841 78063622 Dnmt3l

Change to the required format with:

awk '{print$5","$1":"$3"-"$4","$2}' results.txt | uniq

This gives:

Sox2,chr3:34649994-34652460,+
Rex2,chr4:147021849-147060799,+
Nanog,chr6:122707564-122714633,+
Nanog,chr6:122707488-122714633,+
Nanog,chr6:122707564-122714633,+
Dnmt3l,chr10:78042286-78063622,+
Dnmt3l,chr10:78049958-78063622,+
Dnmt3l,chr10:78055334-78063622,+
Dnmt3l,chr10:78049841-78063622,+

Why are there multiple rows for some genes? Because these genes have more than one transcript

[cfarmeri]

rbagnall , thanks. My genes are UCSC gene symbols, so your suggestion is very suitable.

Answer 3

[MAPK]

There is this tool in Rpackage called: org.Hs.egREFSEQ2EG (https://www.bioconductor.org/packages/3.3/data/annotation/manuals/org.Hs.eg.db/man/org.Hs.eg.db.pdf)

I always use this for various conversions. I have this snippet of R code for you to play around:

 x <- org.Hs.egREFSEQ2EG
    # Get the RefSeq identifier that are mapped to an entrez gene ID
    mapped_seqs <- mappedkeys(x)
    # Convert to a list
    xx <- as.list(x[mapped_seqs])
    if(length(xx) > 0) {
      # Get the entrez gene for the first five Refseqs
      xx[1:5]
      # Get the first one
      xx[[1]]
    }

#Now convert this to dataframe
mydb.refseq<-as.data.frame(x)