ncAA-protein mutageneis scoring

When quantifying the binding energy of a ncAA-protein complex, we observe varying energies for each trial (noise spikes of -0.3 to 0.7 REUs for binding energy & 13 to 21 REUs for folding energy) at a given residue after mutagenesis which doesn’t seem correct given our system.

Our workflow is as follows: we separate a protein-cAA chain, “score” the constituents, mutate the AA to an ncAA & “re-score.” Do you know why we could be experiencing such fluctuations in REUs given the same complex/interface residues? In addition, is there a distinct difference between 2 & 3 or 10 & 20 REUs (same order of magnitude) or can the difference between these be considered negligible where the “trend” is most important given a residue with varying scores when mutated.

The issue names no file, test, or notebook entry point. Begin by reviewing the separation, scoring, and mutation workflow described in the report and determine how repeated scores vary. A useful resolution would explain the fluctuation source and clarify whether the reported REU differences are meaningful or expected.