Entering edit mode
6 weeks ago
Alex
•
0
I have a list of RNA binding proteins from HyDRa prediction. I have a long list of mRNA, I'm interested in.
- For the list of RNA binding proteins, I want to find the potential RNA binding partners. Lets say each protein can bind to 5-20 RNAs, I want to find this list.
- Now for each RNA-protein binding, I want to calculate the binding affinity or any equivalent measurement.
Is there any tools, I can use to do this? or any related insilico approach I can try?
If you asked this 5+ years ago, I would have said NO to both questions. I can still say the same for your #2 as there is no way to reliably predict binding affinity. Pretty sure #1 can't be done by prediction either, but maybe there is some recent DL development I missed.
Thank you for your answer. I'm reading papers on this, still it is hard to find a perfect algorithm.
Unsure how usable this would be, but there is a Protein::DNA/RNA fold and dock approach released from the RoseTTAFold team, paper here. They claim their models self-assessment scores are accurate. You wouldn't get a quantitative binding affinity value, but in theory you could feasibly get a Boolean value of binding to specific mRNA sequences which theoretically could be used for question 1, but you'd have to test it thoroughly.