EvoEF Energy Details

This forum is shown on the index page along with all topics.

Moderator: robpearc

dgong
Posts: 1
Joined: Tue Jan 18, 2022 11:13 pm

EvoEF Energy Details

Post by dgong »

I'm interested using the EvoEF CLI tool to look at some protein structural data, I was hoping to get some clarification on the underlying (biophysical) meaning for some energy function terms:

Why are the reference energies positive? I guess I'm just wondering, how are they obtained/what are they based on?

Are there specific classes for which this tool is more suitable? I'm wondering if the structure I provide must be completed? If, for example, I were to be analyzing just a subunit of a folded protein complex, would the calculated energy value be less meaningful?
xiaoqiah
Posts: 5
Joined: Tue May 04, 2021 5:05 pm

Re: EvoEF Energy Details

Post by xiaoqiah »

Thank you for your interest. Here are my answers to your questions:

Q1: Why are reference energies positive? I guess I'm just wondering, how are they obtained/what are they based on?
A1: Please refer to the main text of this paper (J. Mol. Biol. 2019, 431, 2467-2476) for the overview of parameters including reference energies optimization and Text S3.2 and Text S3.3 of the supporting information for the details of how these parameters are determined. Basically, the reference energies do not have to be positive (e.g. ref_ILE = -0.12, Table S2). The reason that the majority of these ref values are negative because their initial values, together with those for the physical terms, were set to be one. After Monte Carlo optimization, most values still keep a positive value. In fact, the absolute values of the reference energies are not useful but their differences are meaningful. No matter these values are positive or negative, their differences should be identical.

Q2: Are there specific classes for which this tool is more suitable? I'm wondering if the structure I provide must be completed? If, for example, I were to be analyzing just a subunit of a folded protein complex, would the calculated energy value be less meaningful?
A2: EvoEF was optimized and benchmarked on thermodynamic changes (ddG for stability or binding affinity) upon mutations and therefore EvoEF is specifically suitable for this task. You can always use the "ComputeStability" module to calculate the energy for a whole structure but this absolute stability value is meaningless. But the difference between two stability scores upon a mutation (i.e. ddG_stability) is meaningful. For binding energy, it is similar, the binding score calculated by "ComputeBinding" is not calibrated with the experimental binding affinities. However, the difference of two binding scores is useful as you can use it to compare the relative strength of two binding interactions.
Whether you need a complete structure or not depends on your needs. As I mentioned above, the absolute stability score for either a complex or a subunit is meaningless if it is what you want. You should always consider ddG in all scenarios.
Post Reply