strand specificity of TFBS

Question

[Chris Gene]

In querying putative transcription factors that regulated my gene of interest, how important is it do do a strand specific query?

Tx!

[Chris Gene]

Thank you for your reply Michael. I am querying a large dataset of genes for a putative binding site of 6nt (no position matrix motif - I constructed the PWM with 1 as weight for each base in a meme format) with no strand information. My input were the fasta files for each TSS ± 10kb - I downloaded the fasta files using Galaxy, where the input was a bed file with strand information. I am assuming that the fasta file was output according to the strand, in which case there wouldn't be a problem, right?

[Michael]

In case the TF binds both strands(*), then the answer is - unfortunately - yes, you have to do a query on both strands using the reverse complement (of the query, that will be faster). This, even though both, query and reference are directed (gene sequence should be given strand specific).

(*) http://biology.stackexchange.com/questions/30418/do-transcription-factors-bind-to-both-strands-of-dna

Answer 1

[Michael]

Very! You did not specify where you got the motif from and which database you are querying. But unless the search also takes reverse complement into account the motif could be given in the "correct" direction.

These leaves us with two simple questions:

• what's the strand of the motif that is detected? (is it given wrt the transcript or the chromosome?)
• what's the strand of the motif in the DB? (same)

Unless this is specified, and it should be in the documentation of either data set, we cannot say anything more.

After thinking a bit about it, conclude that the search needs to be conducted with both strands in almost all settings.

Answer 2

[mt1022]

Have a look at this article: Lis M, Walther D. The orientation of transcription factor binding site motifs in gene promoter regions: does it matter? BMC Genomics. 2016 Mar 3;17:185. doi: 10.1186/s12864-016-2549-x. PubMed PMID: 26939991.

Our results suggest that for the motifs considered here, either no specific orientation rendering them functional across all their instances exists with orientational requirements instead depending on gene-locus specific additional factors, or that the binding orientation of transcription factors may generally not be relevant, but rather the event of binding itself.

The above cited text is the conclusion of that article. Put it shortly, it seems that the orientation of TFBS does not matter.