# PRODIGY Affinity Scores for Non-native Docking Poses

**URL:** <https://ask.bioexcel.eu/t/prodigy-affinity-scores-for-non-native-docking-poses/5853>\
**Category:** HADDOCK\
**Tags:** haddock\
**Created:** [August 7, 2025, 6:52am UTC](https://ask.bioexcel.eu/t/prodigy-affinity-scores-for-non-native-docking-poses/5853 "2025-08-07T06:52:23Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![George](https://avatars.discourse-cdn.com/v4/letter/g/71e660/32.png) [@George](https://ask.bioexcel.eu/u/George)\
**Post date:** [August 7, 2025, 6:52am UTC](https://ask.bioexcel.eu/t/prodigy-affinity-scores-for-non-native-docking-poses/5853/1 "2025-08-07T06:52:23Z")

</div>

Dear all,

I’m currently working on in silico design of antibody CDRs.  
I generated random CDR sequences, performed docking using HADDOCK to predict antibody-antigen complexes, and estimated the binding affinity using PRODIGY.

Interestingly, several trials consistently yielded PRODIGY ΔG values around -10 kcal/mol, which is close to the experimentally determined ΔG of known complexes.I’m wondering why this might be the case.

Could it be that PRODIGY is not suitable for evaluating binding affinities of complexes generated by docking interfaces that are unlikely to interact in reality?  
In my case, I intentionally specified non-native interface residues to force HADDOCK to generate binding poses.

Thanks in advance.

---

<div class="post-metadata">

**Author:** ![VGPReys](https://dub1.discourse-cdn.com/flex013/user_avatar/ask.bioexcel.eu/vgpreys/32/960_2.png) [@VGPReys](https://ask.bioexcel.eu/u/VGPReys)\
**Post date:** [August 7, 2025, 7:41am UTC](https://ask.bioexcel.eu/t/prodigy-affinity-scores-for-non-native-docking-poses/5853/2 "2025-08-07T07:41:30Z")

</div>

Hi George,

PRODIGY is a supervised contact-based bayesian model, trained on a set \>100 complexes to predict they ΔG values.  
It has not been trained to differentiate true interface from wrong ones.  
Also, it was not trained to predict the ΔG of antibody-antigen complexes.

---

<div class="post-metadata">

**Author:** ![George](https://avatars.discourse-cdn.com/v4/letter/g/71e660/32.png) [@George](https://ask.bioexcel.eu/u/George)\
**Post date:** [August 7, 2025, 8:03am UTC](https://ask.bioexcel.eu/t/prodigy-affinity-scores-for-non-native-docking-poses/5853/3 "2025-08-07T08:03:02Z")

</div>

Thank you very much for your response!

Apologies for the additional question, but as far as I understand, the PRODIGY score is calculated based on the types of residue-residue contacts (e.g., polar/polar, charged/polar, etc.).

So intuitively, I would expect that if a complex is artificially forced into an incorrect binding mode, the resulting contacts would involve a lot of steric clashes or unfavorable interactions, leading to a worse (i.e., less negative) predicted affinity.

Is that not the case?

---

<div class="post-metadata">

**Author:** ![amjjbonvin](https://dub1.discourse-cdn.com/flex013/user_avatar/ask.bioexcel.eu/amjjbonvin/32/23_2.png) [@amjjbonvin](https://ask.bioexcel.eu/u/amjjbonvin)\
**Post date:** [August 7, 2025, 9:06am UTC](https://ask.bioexcel.eu/t/prodigy-affinity-scores-for-non-native-docking-poses/5853/4 "2025-08-07T09:06:29Z")

</div>

HADDOCK does refine the interfaces. So I won’t expect clashes.

And in the flexible refinement residues might try to find favourable interactions…

Not a simple problem.