Thank you for taking the time to read my question.

I have two distinct poses of a protein (poseA.pdb and poseB.pdb) and would like to know the transition state structure (saddle point) between them.

Initially, I would prefer to avoid high computational costs and perform the calculations in a vacuum environment. Assuming the use of a suitable force field such as CHARMM, what would be an appropriate approach to estimate the transition state? If possible, I would appreciate sample code and the computational time (minutes? hours? days? I would like to know the scale).

Additionally, regarding the minimum energy path (MEP) leading to the transition state, methods such as the nudged elastic band (NEB) method and the string method are well known. However, I have seen these methods applied to small molecules rather than high-dimensional systems like proteins.

• Can these methods be effectively applied to proteins?
• If so, what are the key challenges?
• Furthermore, since these methods determine the MEP, does that imply that the transition state structure cannot be directly obtained?

Another approach to finding a saddle point is gradient ascent dynamics. Would this method be suitable for exploring the transition state of a protein? I understand that it may not necessarily find the transition state between the two given poses, but would it still be a viable approach?

I would greatly appreciate any insights, references, or practical guidance on these topics.

Thank you in advance for your time and consideration.