I don't know off the top of my head of a method for the purification of capped RNAs directly - by e.g. affinity purification or anything - but, depending on what you want to do with the RNA, there are methods to selectively generate ligatable RNAs from capped RNAs.
When I was doing the circularised rapid amplification of cDNA ends (cRACE) protocol in my PhD, we would treat total cellular RNA with shrimp alkaline phosphatase (SAP), clear that with phenol, then treat the RNA with tobacco acid pyrophosphatase (TAP); this way, all non-capped RNA in the initial pool was 5' dephosphorylated (and thus non-ligatable) from the first step, but the capped RNAs were protected by the cap; in the second step, the TAP removed the m7G cap, exposing a ligatable 5' phosphate exclusively on once-capped mRNAs. This could then be used for circularisation (or e.g. linker ligation in 5' RACE) to allow exclusive amplification of capped RNAs. You could potentially adapt this kind of protocol by doing a 5'-3' exonuclease step instead of the SAP treatment if you wanted to destroy anything that wasn't capped.
I don't know if this strategy is of any use to you - it really depends on what you want to do with it afterwards - but I thought it might be worth mentioning. Let me know if you have any questions!
Hi Daniel, thanks for the answer. Yes, I need only the mRNA so I would like to eliminate everything trying to preserve, obviously, the quality of the mRNA. I need it for MS analysis so I can't amplify it. Could you explain me better your technique please?