I don't think such a review exists, that said, you can still gather the papers on enhancer predictions methodologies, the most famous paper is chromHMM, but there are some other methodologies based on machine learning techniques like SVM for example. As we don't know yet everything about enhancers in term of functional properties and epigenetic ones, I think it is better to start with all review on enhancers (I guess you did). If you are developing a tool on enhancer predition, my advice is to use experimental data as foreground/background which is in term of data quantity not that much. A lot of tools especially those based on SVM are trainined on a very few amount of experimental data (known enhancers, may be a couple of handreds) which makes it a bit hard to make a solid statement on prediction efficiency. The only way to confirm the accuracy is to do a functional analysis of upstream and downstream regions for an approximation.
6.5 years ago by
Rad • 800