I came across a lot of suggestions to use AlphaFold for 3-D structure prediction of my proteins but I have experienced that I always get better results with trRosetta than AlphaFold and the PROCHECK confirms this.
Hi Mamuna,
The accuracy of protein structure prediction (PSP) methods varies case to case basis. AlphaFold2 is indeed a state of the art method for PSP, so is RoseTTAFold. TrRosetta is another variant of Rosetta. You should use a consensus approach by modeling your protein with different state of the art methods.
TO validate the accuracy of model structure, PROCHECK may not provide you very clear picture. Consider using ModFold8, ProTSAV, ProFitFun, etc.
Thank you Rahul for very informative input. It's always great to learn about new methods and fun to try them. I will definitely try the said servers for next findings.
I agree with @rahul kaushik, both are equally good, and you have to go with consensus output from multiple tools. It depends on the protein which you are analysing.
I would like to enter this discussion. As I was reading within this paper- Article The trRosetta server for fast and accurate protein structure...
that clearly explained the why trRosetta is better than AF2. Please put some light here too.- Badges
- Science topic