Dear colleagues,
I am wondering if nowadays exists some way (ideally webserver/tool) how to predict regulatory elements (or cis-regulatory regions) in my DNA alignment de novo, i.e. based mainly on relative conservancy of DNA alignment regions and other features. This could be theoretically done using neural networks, but I was unsuccessful with my Google search...Maybe someone here has experience with such a task (?).
Thanks for your suggestions,
Martin
Have you tried Evoprinter:
EVOPRINTER, a multigenomic comparative tool for rapid identification of functionally important DNA
DOI: 10.1073/pnas.0506915102
Dear Dr. Uwe Borgmeyer ,
thank you very much for your recommendation, this is exactly what I had in mind ... unfortunately, I am into Arabidopsis and it seems that EVOPRINTER supports only Flaviviruses, Filoviruses, Bacteria, Nematodes, Mosquitoes, Drosophila, Hymenoptera, and Vertebrates.
Best,
Martin
It's technically straightforward to train a neural network to predict "regulatory elements" in silico, based on sequence conservation and other features. But such an approach might not be practically useful because there are some cheap and better experimental approaches there. For example, doing H3K27ac ChIP-seq could easily map all the active enhancers to the genome. Why predict when you can identify them directly?
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics