any good websites for promoter prediction? - (Oct/09/2006 )

Hi all,
When making promoter fusion or complementing constructs, we would all need relevant info to make sure sufficient length of DNA seq is included. So, does anyone of you know certain good webistes which make predictions on the promoter region of a gene? I know some sites do a good job in predicting the -10 and -35 regions--but how about other upstream regions that are to be bound by the transcription factors?

thanks so much for sharing...


Paula

---

Hi Paula,
sorry this is not your answer. I read your post and I thought you may help me. I want to make constructs of promoters fragments (different lenght to evaluate its function). I have no experience at all here, could you please guide me a bit or tell where to start reading?
I have read the invitrogen toporeporter kits (https://catalog.invitrogen.com/index.cfm?fuseaction=viewCatalog.viewProductDetails&sku=&productDescription=707) and it seems I need to transform E coli and then isolate DNA plasmid from correct clones to only then transfect my primary cells. Is that so? I thought I could make the contruct, check it and then transfect directly.
And about the construct, how can I know that the insert have been included in the right orientation?
I will be very grateful for any help!
Karen

---

Hi Paula,

There are loads of prediction servers availiable on the internet. Here are just a few suggestions:

1. Neural Network Promoter Prediction (NNPP) - a method that finds eukaryotic and prokaryotic promoters in a DNA sequence in which a neural network is trained to recognize promoter elements.
Available at http://darwin.nmsu.edu/~molb470/fall2003/P...er_predicti.htm

2. Promoter 2.0 Prediction Server - It builds on principles that are common to neural networks and genetic algorithms, Available at http://www.cbs.dtu.dk/services/Promoter/

3. McPromoter MM:II -- The Markov Chain Promoter Prediction Server, Available at http://genes.mit.edu/McPromoter.html

Hope this helps

Good Luck!

---

www.genomatix.de

see if you can't find their tutorial on promoter-searching and transcription-factor sifting

it's almost impossible to get the whole thing; it can be up to 10kb away, and also on the 3' end and in untranslated regions. I think most people go for the immediate couple of kb and hope for the best; it seems that many relevant sequences are often there, but this isn't a guarantee that you will get everything important.

and I suppose it depends somewhat on how you define a promoter; I would consider a promoter as any contiguous sequence important for regulation of gene transcription

---

I agree--that is why I wish to find a convenient web-based tool to help with predictions of binding sites for various transcription factors.....

thanks for the useful info you all posted--I will definitely check that out carefully...since I am work on bacteria, the tool that works for predicting prokaryotic promoters would be most relavant to me....

cheers,

Paula

---

Hi, karen,
sorry that I do not work on Eukayote system--I may not be familiar with the techniques you work with...so I would not attemp to answer all of your questions, for I do not want to mislead you....


Please help to correct if I am wrong--in my case of making promoter fusion or complementing construct, I tend to use certain web-based tools to check the sequence I am planning to ligate into the vector (lacZ-promoterless reproter plasmid etc). And I pay special attention to the orientation in case for lacZ-fusion. Since I do not have any good luck with ligating PCR product directly, I always ligate the PCR product into a TA cloning vector first and then move to the final vector. As for checking the orientation of the insert, I double check with both sequencing and restriction digestion. And I also hear about using one vector-specific primer and one insert-specific primer for colony-PCR, which may help with checking the insert orientation as well....


May this help a little--and post your questions for further discussions...

I learned a lot from this forum--people are so quick to respond and propose new ideas for trial...

cheers,

paula

---

First I don't think there are programs that can predict anything beyond the core promoter. This is because promoter sequences are much less conserved than coding sequences and there are no common features that define upstream sequences such as enhancer, silencer, TF binding sites, etc. Even for core promoter prediction, programs perform very badly on promoters that don't have a CpG island.

For online programs, my favorite is the Berkeley Neural Network Promoter Prediction

Also this paper gives a very nice comparison and evaluation of all if not all available programs.

Bajic VB, Tan SL, Suzuki Y, Sugano S. Promoter prediction analysis on the whole human genome.
Nat Biotechnol. 2004 Nov;22(11):1467-73.

---

I would like to give you an outline how to clone and characterize a promoter sequence.

1. If the promoter (or 5'-flanking sequence of the gene) sequence is unknown, you have to do a genome walk to find the unknown sequence.

2. If the sequence is known and can be retrieved from databases, you need to design primers that can amplify the promoter sequence for example 1-2 kb upstream the mRNA sequence.

3. Depending on the vector system you are going to use, you may have to add restriction site to the 5' end of your PCR primers so that after PCR, you can digest your product and ligate it into the vector. This way you are sure the insert are in the right orientation. Alternatively, you can do a TA cloning first (TA cloning is easier than the restriction method), then cut the insert from the TA vector with enzymes compatible with your promoter vector (usually luciferase report vector) and subclone the insert into the report vector.

4. Do serial deletion using the longest promoter construct as a template to obtain constructs of different sizes.

5. To obtain plasmid DNA for transfection into cultured cells, you need to transform the constructs into competent E. Coli cells, then do a mini plasmid preparation. The amount of plasmid from miniprep should be good enough for quite a few transfections.

Good luck.

---