[BioC] reverse complement or no reverse complemnt on biomaRt / biomart.org

[James W. MacDonald]

Tefina Paloma wrote:
> 
> I apologize for giving such a vague description.
> 
> If I go to the ensemble page, click on "Mine Ensembl with BioMart", chose as
> database ensembl 56, Homo Sapiens, GRCh37, filter by ENST00000280193, click on
> results, and then click on the ENST00000280193 which leads to the transcript
> summary of this ensembl. Then, on the left hand side, I click on "export data"
> and I chose "fasta format", "feature strand", "flanking sequence" and "5utr".
> By doing this, I get the sequences I posted.

That looks like a bug in that export wizard. It doesn't matter which 
strand you choose for the 5' UTR, you get the (correct) reverse strand 
regardless.

However, if you choose either feature strand or reverse strand, you get 
the forward strand (the reverse complement of the correct 5' flanking 
region). If you choose forward strand, you still get the forward strand, 
but it is off the 3' end of the gene.

Try pasting this into blat (http://genome.ucsc.edu/cgi-bin/hgBlat) to 
see what I mean.

 > biomaRt
CCGCCGCCAGCGCCCCCGCCGCAGCGCCCGCGGCCCGGCTCCTCTCACTT
 > forward
GTAGTTTTCGTTCATCATGTAAGATGATAATGGACTGAACTTAGCAGCTT
 > feature
AAGTGAGAGGAGCCGGGCCGCGGGCGCTGCGGCGGGGGCGCTGGCGGCGG

So the only correct results here are from biomaRt.

Best,

Jim




> 
> Best,
> Tefina
> 
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor

-- 
James W. MacDonald, M.S.
Biostatistician
Douglas Lab
University of Michigan
Department of Human Genetics
5912 Buhl
1241 E. Catherine St.
Ann Arbor MI 48109-5618
734-615-7826