bed file around the TSSs

[joseph]

Hi
I would highly appreciate if you could show me how to create a bed file around the TSSs from UCSC databases such as ensembl or refSeq genes. I need 350 nucleotides upstream and 150 nucleotides downstream of TSSs. The bed file should look like below, where:

chromStart is 350 nucleotides upstream of TSS
chromEnd is 150 nucleotides downstream of TSS
name is Name of gene or transcript_id depending on the database.

chrom	chromStart	chromEnd	name		score	strand
chr1	67051159	67163158	NM_024763	0	-	
chr1	67075869	67163158	NM_207014	0	+	
chr1	16762998	16812569	NM_017940	0	-

Thanks
Joseph

[malachig]

Have you considered using the EnsEMBL API for these kinds of tasks?

Ensembl API docs

[adamdeluca]

On the UCSC table browser go to the "Genes and Gene Predictions" group, select the table you want (ensGene, refGene, etc). Change the output format to "Selected fields..." and on the next page select the chr and the TSS. You can then pad the TSS by doing something like:

awk '{print $1"\t"$2-349"\t"$2+350}' file > TSS.bed