For all bioinformatic experts, I have a known sequence (DNA) and I'd like to find the DNA-binding consensus seq. between my known seq. of interest and possible others (unknown) targets. Which is the best tool to use in order to enter my known seq. and find possible targets for binding? In other words I want to find possible targets for my protein of interest for protein-protein interactions.Thank you!
The first part of your question makes it sound like you have a DNA sequence that you believe is bound by a (known?) DNA-binding protein, and you would like to figure out
1)What specific portion of your DNA sequence is bound (the binding site),
and
2) What other sequences in your genome are also bound?
But then, you mention targets of your protein of interest, for protein-protein interactions.
Could you provide a bit more information?
Thank you for the reply and I am sorry for the confusion. I am looking for possible targets for my known sequence. I know that my seq. of interest is a transcriptional factor, that means it binds on dna of other genes. So I am searching for these possible targets and find possible binding sites between my known seq. and those targets so I could test them at bench. Did I clear it out?
If you don't have information about the specific sequence motif bound by your transcription factor, you have to perform additional experiments. For instance ChIP-Seq using antibodies recognizing your transcription factor, then get the regions where the TF binds the more, and proceed from point (ii) below. Alternatively you may have expression data in a (or better more) conditions where your transcription factor is active and changes in activity, better if you have gene expression data in a null mutant of your TF if you can. If you have gene expression data in a mutant of your TF (null or constitutively active), then you could for instance (i) identify differentially expressed genes; (ii) extract upstream sequences for those genes, (iii) run tools like AlignAce or MEME (or others available) to identify enriched sequence motifs, and if you find some motif common to most or all the genes changing gene expression, than you can proceed and try to perform focused experiments to see if the identified sequence motif is indeed bound by your TF. Another alternative is if you don't have a mutant of your TF but you have e.g. a gene expression dataset with multiple conditions. In that case you could run some gene regulatory network inference algorithm (for instance you could use the MINET package in R-cran or similar) to see if you get a group of genes with a gene expression profile that is highly correlated with the TF of interest, and proceed again from the above point (iii). If you have some previous knowledge about one or a few targets you can immediately try with DNAase I mapping or things like that to see where your TF exactly binds. All of this might work pretty well in Bacteria or yeast (and other "lower" eucaryotes), but it is much more difficult in human and plants, where promoters span large regions, because the tools for finding enriched sequence motifs work better if the ratio among the length of the enriched motif and the length of the sequence used as input is not too small. Sometimes the information about the sequence bound by a TF is known and you can find it in some database depending on the organism you are considering. In that case you can use the model of the sequence motif to look for additional occurrences upstream of other genes.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics
- Bioinformatic Tools