TargetScan v6.1 genome coordinates

jmw86069 · 05-09-2012, 09:21 AM

I'm trying to find or create genome coordinates for TargetScan v6.1, released March 2012. UCSC provides a track from v5.1, but nothing newer. This thread (link UCSC mailing list) suggests it's trickier than one would think. So before embarking on that quest, I wanted to poll this board to see if anyone already has created genome coordinates from TargetScan v6.1?

Here is my best guess workflow:

Download UTR_Sequences.txt from TargetScan for the given species, grep out only the taxon ID needed (10090 for mouse from the TargetScanMouse data; 9606 for human from the TargetScanHuman data; etc.)
Create gapped-to-ungapped mapping of the UTR_Sequences.txt data (which is in gapped alignment format) so I can convert the reported TargetScan gapped coordinates to un-gapped UTR coordinates.
Strip out the gaps from each sequence to create an un-gapped UTR FASTA file, then BLAT them to the appropriate genome (mm9, hg18/hg19, etc.), in order to convert UTR ungapped coordinates to genome coordinates. This step seems necessary because the TargetScan UTRs were created using some logic, and aren't directly from any available UCSC track (that I can tell anyway.)
Use the Conserved_Family_Info.txt file for the given organism, which provides gapped UTR coordinates for each predicted miRNA site. Convert each coordinate using the logic above.
(gapped UTR coordinate) --> (ungapped UTR coordinate) --> (genome coordinate)

I don't know how much I lose by using BLAT to re-align UTRs back to the genome. If anyone has a suggested alternative (e.g. MUMmer?), I'd greatly welcome it! I'm really hoping for well-behaved alignments!

DrOM · 07-02-2013, 05:01 AM

Hallo!

have you advanced with the genomic coordinates for TargetScan6.2? I think you can skip step #2, since the target file also provides the position in the ungapped UTRs of that species. Therefore it should be ok to strip the gaps of the alignment file, get the genomic coordinates for the UTRs and proceed like you outlined above.

It would be great if you could share your results!!
Cheers

Pablo

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
TargetScan input	NicoBxl	Bioinformatics	3	05-13-2014 06:01 AM
How to convert genome coordinates from two assemblies	TuA	Bioinformatics	5	11-20-2013 07:43 PM
How to align SNP coordinates against a circular reference genome?	Lspoor	General	1	02-17-2012 10:44 AM
coordinates in alignment	m_elena_bioinfo	Bioinformatics	1	01-21-2011 06:07 AM
targetscan files	NicoBxl	Bioinformatics	0	12-13-2010 05:38 AM

05-09-2012, 09:21 AM	#1
jmw86069 Member Location: RTP, NC, USA Join Date: Jun 2009 Posts: 28	TargetScan v6.1 genome coordinates I'm trying to find or create genome coordinates for TargetScan v6.1, released March 2012. UCSC provides a track from v5.1, but nothing newer. This thread (link UCSC mailing list) suggests it's trickier than one would think. So before embarking on that quest, I wanted to poll this board to see if anyone already has created genome coordinates from TargetScan v6.1? Here is my best guess workflow: Download UTR_Sequences.txt from TargetScan for the given species, grep out only the taxon ID needed (10090 for mouse from the TargetScanMouse data; 9606 for human from the TargetScanHuman data; etc.) Create gapped-to-ungapped mapping of the UTR_Sequences.txt data (which is in gapped alignment format) so I can convert the reported TargetScan gapped coordinates to un-gapped UTR coordinates. Strip out the gaps from each sequence to create an un-gapped UTR FASTA file, then BLAT them to the appropriate genome (mm9, hg18/hg19, etc.), in order to convert UTR ungapped coordinates to genome coordinates. This step seems necessary because the TargetScan UTRs were created using some logic, and aren't directly from any available UCSC track (that I can tell anyway.) Use the Conserved_Family_Info.txt file for the given organism, which provides gapped UTR coordinates for each predicted miRNA site. Convert each coordinate using the logic above. (gapped UTR coordinate) --> (ungapped UTR coordinate) --> (genome coordinate) I don't know how much I lose by using BLAT to re-align UTRs back to the genome. If anyone has a suggested alternative (e.g. MUMmer?), I'd greatly welcome it! I'm really hoping for well-behaved alignments!

07-02-2013, 05:01 AM	#2
DrOM Junior Member Location: Germany Join Date: Jul 2012 Posts: 5	Hallo! have you advanced with the genomic coordinates for TargetScan6.2? I think you can skip step #2, since the target file also provides the position in the ungapped UTRs of that species. Therefore it should be ok to strip the gaps of the alignment file, get the genomic coordinates for the UTRs and proceed like you outlined above. It would be great if you could share your results!! Cheers Pablo