My current research is about a gene transcription regulation.
I try to find its promoter region and the putative transcription factor-binding elements.
After I found out the upstream of 5' flanking region of my interested gene (About 1Kb), I used the "WWW Promoter Scan " to search the possible TFBs.
Today I find another software of promoter scanning , "MatInspector promoter analysis tool"
After I paste the 1Kb sequence, there are many TFBs that I don't know how to select possible TFBs
I need some of your recommendations ~XD
Try this: (SMART software) from Genome-wide transcription factor binding site/promoter databases for the analysis of gene sets and co-occurrence of transcription factor binding motifs.
MatInspector is nice, but there are approaches better than that. Let's say, first, that all but 1 tools presently available in the internet are based on positional weight matrices (PWMs), with all their drawbacks. The 1 tools that is alternative to that is SiTaR, see E. Fazius et al, 2011, Bioinformatics, and https://sbi.hki-jena.de/sitar/index.php . The main problem in TFBS detection and prediction is the number of false positives (that is what makes the number of predictions per promoter so hard to interpret). Last year we made a comparison of several best-presenting PWM-based tools in terms of the number of false positive predictions they provide and concluded that the best in this respect are the searching tool of Jaspar database (http://jaspar.genereg.net/cgi-bin/jaspar_db.pl) and P-Match (http://www.gene-regulation.com/pub/programs.html#pmatch). The problem with Jaspar is that it has not that many TFs included, but those that are there are realiable; P.Match has an outdated TFBS library. The SiTaR tool tends to outperform both those PWM based tools, but to use it you have to collect somewhere the sets of TFBSs for each of TFs you're interested in. For instance, in Genomatix, if you have licence for that, or in Transfac (also needs licence). Otherwise, you can get the motifs from Jaspar, but probably they are available not for all TFs.
In any case (independently on the tool you apply), the number of predicted TFBSs for your promoter will depend on the searching parameters. The rest is the play with those parameters and cutting, in such way, the weaker hits, retaining the strongest. In such way, you can run a risk to lose something useful, but normally for the first glance it works fine.
I wish you good luck! In case you have further questions, just ask!
P.S. To further refine your search you can use the tool DistanceScan, which looks for pairwise combinations of TFBSs.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics
- Bioinformatic Tools