PicTar is an algorithm for the identification of microRNA targets. PicTar is a project of the Rajewsky lab at NYU's Center for Comparative Functional Genomics and the MDC, Berlin.
The PicTar website provides details (3' UTR alignments with predicted sites, links to various public databases etc) regarding:
microRNA target predictions in vertebrates (Krek et al, Nature Genetics 37:495-500 (2005))
microRNA target predictions in seven Drosophila species (Grün et al, PLoS Comp. Biol. 1:e13 (2005))
microRNA targets in three nematode species (Lall et al, Current Biology 16, 1-12 (2006))
human microRNA targets that are not conserved but co-expressed (i.e. the microRNA and mRNA are expressed in the same tissue) (Chen and Rajewsky, Nat Genet 38, 1452-1456 (2006))
UPDATED PICTAR PREDICTIONS 2012: doRiNA: a database of RNA interactions in post-transcriptional regulation (Anders et al, Nucleic Acids Res. 2012 Jan;40(Database issue):D180-6. Epub 2011 Nov 15.)
PicTar miRNA target site predictions for the hg17, mm7, dm2 and ce2 genomes can be obtained from the UCSC genome browser via the 'tables' feature.
New predictions for hg18, mm9 and ce6 are available on our own UCSC mirror at http://dorina.mdc-berlin.de
For your convenience, doRiNA offers a link for downloading PicTar target site predictions: simply run a doRiNA query for all miRNA targets, setting the search options to 100% under step 3, and follow the link on the top of the results page.
To build contiguous alignments for each 3' UTR or other types of mRNAs (5' UTRs, CDS, ..) from the MultiZ multiple alignments at the UCSC Genome Database you can use the scripts from AlignmentBuilder.
- The format of the gene annotation table available from UCSC changes occasionally. Check that the strand is given as "+" or "-", and not just "+" or blank. In the latter case you need to insert the missing "-"
- The program also requires that the gene annotation table have exactly 10 fields - no "bin" field at the beginning and no extra fields at the end. Currently, there is a short script, transform_annotation.py that removes the first field and the extra fields at the end, but watch this in the future in case the format changes again.
- There is an off-by-one error with the multiZ alignments file. You need to add 1 to the second field of that file or the alignments will be off by one.
- Do not use the characters "." or "-" in the gene name because this will cause errors.