Figure 3S. Amino acid composition analysis of the HDAC8 binding pocket

Residue-type frequency within Pocket 1 (catalytic site, 2020 Å³) for all 15 compounds, the relationship between aspartate contacts and binding score, and the aromatic/non-aromatic balance of the engaged residues.

(A) Top-5 residue types per compound

Counts pooled across both chains; ordered by Pocket 1 binding score.

(B) Aspartate count vs binding score

Novel derivatives (n = 9): r = −0.82, p < 0.01 · all 15: r = −0.09 (ns)

Aromatic = PHE + HIS + TYR · Non-aromatic = ASP + LYS + GLY + PRO · (ref) = reference compound.

Residue composition data (Pocket 1)

Observations

Aspartate and affinity: among the nine novel derivatives, ASP contact count correlates strongly with binding score (r = −0.82, p < 0.01); KB-282 shows the highest ASP count (8) and the strongest binding (−12.3 kcal/mol).
Whole-set behaviour: across all 15 compounds, including references, the ASP–score association is negligible (r = −0.09), indicating that residue count alone does not determine affinity when chemically diverse scaffolds are pooled.
Phenylalanine conservation: six PHE contacts appear for 13 of 15 compounds (KB-269 shows 5, KB-283 shows 4), consistent with a recurring role for aromatic stacking.
Aromatic environment: high-affinity compounds engage 9–14 aromatic residues (PHE + HIS + TYR), supporting π-mediated stabilisation within the catalytic pocket.
Interpretation: the absence of a single-residue linear predictor across the full set points to multifactorial determinants, geometry and interaction type alongside residue identity.

Sources. Residue counts: Supplementary S3 (Pocket 1 composition) and S5 (full residue listings). Binding scores: Table 2 and Supplementary S1–S2. Correlations computed by Pearson coefficient (α = 0.05) directly from the values shown. Colour. Residue palette follows the Okabe–Ito colourblind-safe scheme; point colour in (B) encodes binding tier and hollow markers denote reference compounds.