Genome KpATCC43816

Protein target profile

Putative UDP-glucoronosyl and UDP-glucosyl transferase

Accession: VK055_5103

Gene: AIK83629.1 iroB 3D evidence: AlphaFold DB model + ColabFold model Metabolism Not in network UniProt A4GZE8

Export view

Length 355

Pocket druggability (P2Rank) 0.982

Direct ligand evidence 0 55 total records

Functional annotation 0 EC 3 GO

Target summary

Promising target candidate with multiple supporting evidence streams.

Automated synthesis of the evidence currently loaded. Review the underlying records before prioritizing this protein.

Main supporting evidence

Risks to review

Evidence coverage

Core evidence streams are available for this target.

Terms and data sources used on this page

PDB: experimentally determined structures from the Protein Data Bank. These are the strongest structural evidence, but may cover only part of the protein.

AlphaFold DB model: a precomputed predicted structure downloaded from AlphaFold Database/UniProt, not an experiment performed here.

ColabFold model: a predicted structure generated for this workspace; interpret it with coverage and confidence.

pLDDT: confidence score for predicted structures. High values support local geometry; low values mean the region should not drive pocket interpretation.

FPocket / P2Rank: software tools that predict possible ligand-binding pockets on a 3D structure. They are useful screening signals, not experimental validation.

Druggability: a pocket-based estimate of whether a small molecule could bind productively. It does not mean a drug already exists.

PDB ligand: a compound observed in an experimental structure. Direct same-protein records are stronger than homolog-transferred records.

ChEMBL: a public database of measured compound bioactivity. Direct entries are stronger than entries transferred from similar proteins.

ZINC: a purchasable-compound database. Here it marks proposed candidates from chemical similarity, not measured binders.

LigQ / LigQ_2: an internal Target pipeline step that gathers PDB, ChEMBL, and ZINC ligand evidence for each protein.

Off-target: sequence similarity to proteins we prefer not to hit, such as human proteins or beneficial gut microbiome proteins.

DEG: Database of Essential Genes. A match suggests the protein resembles genes known to be essential in other organisms.

Roary / CoreCruncher: pan-genome tools used to decide whether a gene is core across analyzed strains or accessory/strain-specific.

EC / GO: functional annotations: EC describes enzyme reactions; GO describes biological process, molecular function, or cellular component.

KEGG pathway: a curated metabolic route label used here to group reactions imported from the metabolic model.

Chokepoint: a metabolic reaction that is the only producer or consumer of a metabolite in the imported model.

Prioritization evidence

Selectivity, essentiality, structural confidence, conservation, and predicted binding-site evidence.

Off-target risk

Human off-target: No hit
Gut microbiome similarity: 0.2% of screened genomes Lower prevalence suggests narrower overlap with the screened gut microbiome.

Essentiality

Essential (DEG): Y
DEG identity (%): 67.797 Higher values support similarity to known essential genes.
DEG E-value: 1.88e-178 Smaller values mean stronger essential-gene similarity.

Structure confidence

ColabFold pLDDT: 88.56 0-100 confidence; >70 supports local structural interpretation.

Binding-site evidence

AlphaFold DB / UniProt model

P2Rank's binding-site probability is the primary druggability signal shown across the app; FPocket's druggability score is shown alongside it for comparison. Both estimate small-molecule pocket quality after applying the curated structure priority — neither is experimental binding evidence. The 3D viewer may show a different loaded structure, so visible pockets can differ.

Druggability (P2Rank) 0.982

Structure A4GZE8

Pocket Pocket 1

Druggability (FPocket) 0.979

Structure A4GZE8

Pocket Pocket 16

ColabFold model

P2Rank 0.941 · Pocket 1

FPocket 0.701 · Pocket 26

Core conservation Accessory gene

Roary accessory

CoreCruncher accessory

Gut microbiome 8 / 4744 genomes with a hit

Prevalence 0.2%

Cross-references

External database identifiers for this protein, its structures, ligands, and metabolic reactions.

Sequence

UniProt A4GZE8

Metabolic context

Reactions catalyzed, pathway membership, and centrality in the genome-scale metabolic network.

This protein is not associated with the imported metabolic network for this genome.

Browse the genome's metabolic network

Imported from KpATCC43816.sbml · 2026-07-09

Sequence

Primary amino-acid sequence viewer.

MLSLAQAFRVNGHEVLIASGGTFVKKAAEAGLVAFDAAPGLDSEAGYCRHEALRKESHIGTKMGNFSFFSEEMADHLVEFAGHWRPDLIIYPPLGVIGPLIAAKYDIPVVMQTVGFGHTPWHIKGVTRSLADAYHRHGVGAAPRDMAWLDVTPPSMSILENDGEPIIPMQYVPYNGGAVWEPWWERKPDRKRLLVSLGTVKPMVDGLELITWVMDSASEVDAEIILHLSANARSELRSLPSNVRLVDWIPMGAFLNGADGFIHHGGAGNTLTALHAGIPQIVFGQGADRPVNARVVVERGCGIIPGDGGLSSNMINAFLDNRALRDASEEVAAEMAAQPCPSEVAKKLIAMVQKG

Functional annotations

Enzyme classification and Gene Ontology terms linked to this protein.

3 GO

Subcellular localization

Localization: Cytoplasmic

Gene Ontology (GO)

GO:0008194 Catalysis of the transfer of a glycosyl group from a UDP-sugar to a small hydrophobic molecule.
GO:0016758 Catalysis of the transfer of a hexosyl group from one compound (donor) to another (acceptor).
GO:0017000 The chemical reactions and pathways resulting in the formation of an antibiotic, a substance produced by or derived from certain fungi, bacteria, and other organisms, that can destroy or inhibit the growth of other microorganisms.

Sequence domains and features

Domain and signature matches imported from InterPro and related databases.

8 records

Show feature table

Start	End	DB	Term	Name
181	336	Gene3D	G3DSA:3.40.50.2000	Glycogen Phosphorylase B;
150	244	Pfam	PF06722	Protein of unknown function (DUF1205)
150	244	InterPro	IPR010610	Erythromycin biosynthesis protein CIII-like, central
1	345	Gene3D	G3DSA:3.40.50.2000	Glycogen Phosphorylase B;
183	337	PANTHER	PTHR48050	STEROL 3-BETA-GLUCOSYLTRANSFERASE
1	354	CDD	cd03784	GT1_Gtf-like
1	354	InterPro	IPR002213	UDP-glucuronosyl/UDP-glucosyltransferase
1	352	SUPERFAMILY	SSF53756	UDP-Glycosyltransferase/glycogen phosphorylase

3D structure

Selected loaded structure. Experimental PDB entries may cover only a portion of the sequence; AlphaFold DB and ColabFold models typically cover the full protein but remain computational predictions.

Download VMD script Full viewer

Loading 3D structure...

Drag to rotate — click the view, then scroll to zoom.

Pocket score High Medium Low

How colors and pocket overlays are used

Uniform protein color marks the displayed model as a single molecular object.

Experimental PDB structures may be colored by chain to distinguish subunits or copies present in the file.

Pocket colors and alpha spheres are evidence overlays for predicted binding cavities; they are not alternative protein chains.

'Alpha spheres' is FPocket's own cavity-shape geometry, imported when available and aligned with the loaded structure.

'Pocket atoms'/'Predicted site atoms' show the pocket's residue atoms instead: P2Rank reports residues rather than alpha spheres, and FPocket falls back to this when alpha-sphere geometry is unavailable or doesn't align.

'No pocket geometry' means neither alpha spheres nor residue-position data could be found for that pocket; the layer just highlights the same residues as 'Nearby residues'.

Pocket details Inspect a specific pocket, or open the full viewer

Binding pockets · P2Rank

Probability: high ≥ 0.5 · medium 0.2–0.49 · low < 0.2

Pocket 1 P2Rank #1

0.982

Likely same site as FPocket 16 1.3 Å 34 shared residues 85% of smaller site

Show in viewer

Surrounding area

Pocket 2 P2Rank #2

0.104

Show in viewer

Surrounding area

Pocket 3 P2Rank #3

0.043

Show in viewer

Surrounding area

Binding pockets · FPocket

Druggability: high ≥ 0.7 · medium 0.4–0.69 · low < 0.4

Pocket 1 FPocket #16

0.979 Unusual size

Likely same site as P2Rank 1 1.3 Å 34 shared residues 85% of smaller site

Show in viewer

Surrounding area

Binding pockets · P2Rank

Probability: high ≥ 0.5 · medium 0.2–0.49 · low < 0.2

Pocket 1 P2Rank #1

0.941

Likely same site as FPocket 26 2.7 Å 24 shared residues 100% of smaller site

Show in viewer

Surrounding area

Pocket 2 P2Rank #2

0.586

Show in viewer

Surrounding area

Pocket 3 P2Rank #3

0.489

Likely same site as FPocket 30 6.9 Å 10 shared residues 67% of smaller site

Show in viewer

Surrounding area

Pocket 4 P2Rank #4

0.283

Show in viewer

Surrounding area

Pocket 5 P2Rank #5

0.274

Show in viewer

Surrounding area

Binding pockets · FPocket

Druggability: high ≥ 0.7 · medium 0.4–0.69 · low < 0.4

Pocket 1 FPocket #26

0.701 Unusual size

Likely same site as P2Rank 1 2.7 Å 24 shared residues 100% of smaller site

Show in viewer

Surrounding area

Pocket 2 FPocket #30

0.384

Likely same site as P2Rank 3 6.9 Å 10 shared residues 67% of smaller site

Show in viewer

Surrounding area

All structural evidence 0 experimental · 2 predicted

Structural evidence

0 + 2

Experimental PDB entries plus predicted AlphaFold DB or ColabFold models. Click Switch to display a different loaded structure in the viewer.

Entry	Method	Resolution	Chain	Coverage	Links	Status
AlphaFold DB AF_A4GZE8	AlphaFold DB	—	—	full sequence	—	Viewing
ColabFold VK055_5103	ColabFold	—	—	full sequence	—	Loaded

Ligand evidence

Ligands grouped by evidence source. PDB ligands keep the source crystal visible, and loaded crystals can be opened directly in the structure viewer.

55 records

Chemistry signal

Structural ligand evidence is available for this target.

Direct evidence 0 same-protein records

Transferred evidence 5 records from similar proteins

Structural ligands 5 0 loaded crystals

Measured bioactivity 0 direct and transferred ChEMBL records

Proposed compounds 50 similarity-based ZINC candidates

Best available ligand signal

3R2 PDB via homolog 530.4 Da · LogP -0.74 · TPSA 227.1 Open detail RCSB PDB

C0T PDB via homolog Detail RCSB PDB

DUD PDB via homolog Detail RCSB PDB

PE4 PDB via homolog Detail RCSB PDB

TYD PDB via homolog Detail RCSB PDB

Highest-confidence structural evidence: ligands co-crystallized with this exact protein. If the source PDB is loaded in Target, use Open crystal to inspect it in the structure viewer.

No PDB structure with a co-crystallized ligand found for this exact protein.

Structural evidence inferred from similar proteins. The source crystal indicates where the ligand was observed; the UniProt column identifies the homologous protein carrying that ligand.

Ligand	Source crystal	UniProt (homolog)	MW · LogP · TPSA	Lipinski	PAINS	SMILES
3R2 RCSB PDB	4RIF	Q9ZGC0	530.4 Da LogP -0.74 TPSA 227.1	3 viol.	✓ Clean	`C[C@H]1C[C@@H](C[C@H]([C@@H]1O)O)OP(=O)(O)OP(=O…`
C0T RCSB PDB	3OTI	Q8KND7	584.7 Da LogP 0.66 TPSA 166.8	3 viol.	✓ Clean	`CC1C(C(C(C(O1)OC2C#CC=CC#CC3(CC(=O)C(=C2C3=CCSS…`
DUD RCSB PDB	4AMB	Q9RN61	388.2 Da LogP -1.59 TPSA 197.6	✓ Ro5	✓ Clean	`C1[C@@H]([C@H](O[C@H]1N2C=CC(=O)NC2=O)CO[P@](=O…`
PE4 RCSB PDB	3D0R	Q8KND7	354.4 Da LogP 0.11 TPSA 84.8	✓ Ro5	✓ Clean	`CCOCCOCCOCCOCCOCCOCCOCCO`
TYD RCSB PDB	4G2T	D6MSX4	402.2 Da LogP -1.28 TPSA 197.6	✓ Ro5	✓ Clean	`CC1=CN(C(=O)NC1=O)[C@H]2C[C@@H]([C@H](O2)CO[P@]…`

Experimental bioactivity from ChEMBL measured directly on this protein. Score = pchembl (−log Ki/IC₅₀; higher = more potent).

No ChEMBL bioactivity data found for this exact protein.

Bioactivity inferred from similar proteins in ChEMBL. Score = pchembl (−log Ki/IC₅₀; higher = more potent).

No ChEMBL hits found through similar proteins.

Proposed virtual-screening candidates from ZINC. Score = Tanimoto similarity to a known binder (0–1; higher = more similar).

Ligand	Tanimoto	MW · LogP · TPSA	Lipinski	PAINS	SMILES
ZINC33979243 ZINC	1.000	402.2 Da LogP -1.28 TPSA 197.6	✓ Ro5	✓ Clean	`Cc1cn([C@@H]2C[C@H](O)[C@H](CO[P@@](=O)(O)OP(=O…`
ZINC5650743 ZINC	1.000	222.3 Da LogP 0.07 TPSA 57.2	✓ Ro5	✓ Clean	`CCOCCOCCOCCOCCO`
ZINC6403917 ZINC	1.000	354.4 Da LogP 0.11 TPSA 84.8	✓ Ro5	✓ Clean	`CCOCCOCCOCCOCCOCCOCCOCCO`
ZINC8217150 ZINC	1.000	388.2 Da LogP -1.59 TPSA 197.6	✓ Ro5	✓ Clean	`O=c1ccn([C@H]2C[C@H](O)[C@@H](CO[P@@](=O)(O)OP(…`
ZINC12501922 ZINC	0.958	468.1 Da LogP -1.47 TPSA 244.1	2 viol.	✓ Clean	`O=c1ccn([C@@H]2C[C@H](O)[C@@H](CO[P@@](=O)(O)O[…`
ZINC13436738 ZINC	0.958	468.1 Da LogP -1.47 TPSA 244.1	2 viol.	✓ Clean	`O=c1ccn([C@@H]2C[C@H](O)[C@H](CO[P@@](=O)(O)O[P…`
ZINC13436741 ZINC	0.958	468.1 Da LogP -1.47 TPSA 244.1	2 viol.	✓ Clean	`O=c1ccn([C@H]2C[C@H](O)[C@H](CO[P@@](=O)(O)O[P@…`
ZINC13436743 ZINC	0.958	468.1 Da LogP -1.47 TPSA 244.1	2 viol.	✓ Clean	`O=c1ccn([C@@H]2C[C@@H](O)[C@H](CO[P@@](=O)(O)O[…`
ZINC13436745 ZINC	0.958	468.1 Da LogP -1.47 TPSA 244.1	2 viol.	✓ Clean	`O=c1ccn([C@H]2C[C@@H](O)[C@H](CO[P@@](=O)(O)O[P…`
ZINC13514392 ZINC	0.833	308.2 Da LogP -1.71 TPSA 151.1	✓ Ro5	✓ Clean	`O=c1ccn([C@@H]2C[C@@H](O)[C@@H](COP(=O)(O)O)O2)…`
ZINC1532629 ZINC	0.833	308.2 Da LogP -1.71 TPSA 151.1	✓ Ro5	✓ Clean	`O=c1ccn([C@@H]2C[C@@H](O)[C@H](COP(=O)(O)O)O2)c…`
ZINC3869888 ZINC	0.833	308.2 Da LogP -1.71 TPSA 151.1	✓ Ro5	✓ Clean	`O=c1ccn([C@@H]2C[C@H](O)[C@H](COP(=O)(O)O)O2)c(…`
ZINC3869889 ZINC	0.833	308.2 Da LogP -1.71 TPSA 151.1	✓ Ro5	✓ Clean	`O=c1ccn([C@H]2C[C@H](O)[C@H](COP(=O)(O)O)O2)c(=…`
ZINC3869890 ZINC	0.833	308.2 Da LogP -1.71 TPSA 151.1	✓ Ro5	✓ Clean	`O=c1ccn([C@H]2C[C@@H](O)[C@H](COP(=O)(O)O)O2)c(…`
ZINC4228260 ZINC	0.833	308.2 Da LogP -1.71 TPSA 151.1	✓ Ro5	✓ Clean	`O=c1ccn([C@H]2C[C@H](O)[C@@H](COP(=O)(O)O)O2)c(…`
ZINC6524723 ZINC	0.833	308.2 Da LogP -1.71 TPSA 151.1	✓ Ro5	✓ Clean	`O=c1ccn([C@@H]2C[C@H](O)[C@@H](COP(=O)(O)O)O2)c…`
ZINC35636069 ZINC	0.717	404.2 Da LogP -2.62 TPSA 217.8	1 viol.	✓ Clean	`O=c1ccn([C@@H]2O[C@H](CO[P@](=O)(O)OP(=O)(O)O)[…`
ZINC40655887 ZINC	0.717	404.2 Da LogP -2.62 TPSA 217.8	1 viol.	✓ Clean	`O=c1ccn([C@@H]2O[C@H](CO[P@](=O)(O)OP(=O)(O)O)[…`
ZINC40655889 ZINC	0.717	404.2 Da LogP -2.62 TPSA 217.8	1 viol.	✓ Clean	`O=c1ccn([C@@H]2O[C@H](CO[P@@](=O)(O)OP(=O)(O)O)…`
ZINC44460318 ZINC	0.717	404.2 Da LogP -2.62 TPSA 217.8	1 viol.	✓ Clean	`O=c1ccn([C@H]2O[C@@H](CO[P@@](=O)(O)OP(=O)(O)O)…`
ZINC4490939 ZINC	0.717	404.2 Da LogP -2.62 TPSA 217.8	1 viol.	✓ Clean	`O=c1ccn([C@@H]2O[C@H](CO[P@@](=O)(O)OP(=O)(O)O)…`
ZINC8585026 ZINC	0.717	404.2 Da LogP -2.62 TPSA 217.8	1 viol.	✓ Clean	`O=c1ccn([C@H]2O[C@@H](CO[P@@](=O)(O)OP(=O)(O)O)…`
ZINC8585028 ZINC	0.717	404.2 Da LogP -2.62 TPSA 217.8	1 viol.	✓ Clean	`O=c1ccn([C@@H]2O[C@@H](CO[P@@](=O)(O)OP(=O)(O)O…`
ZINC8585030 ZINC	0.717	404.2 Da LogP -2.62 TPSA 217.8	1 viol.	✓ Clean	`O=c1ccn([C@H]2O[C@@H](CO[P@@](=O)(O)OP(=O)(O)O)…`
ZINC8585032 ZINC	0.717	404.2 Da LogP -2.62 TPSA 217.8	1 viol.	✓ Clean	`O=c1ccn([C@@H]2O[C@@H](CO[P@@](=O)(O)OP(=O)(O)O…`
ZINC12503053 ZINC	0.691	402.2 Da LogP -1.28 TPSA 197.6	✓ Ro5	✓ Clean	`Cc1cn([C@@H]2C[C@H](O)[C@@H](CO[P@@](=O)(O)OP(=…`
ZINC12959005 ZINC	0.691	484.1 Da LogP -2.50 TPSA 264.4	2 viol.	✓ Clean	`O=c1ccn([C@@H]2O[C@H](CO[P@](=O)(O)O[P@](=O)(O)…`
ZINC12959016 ZINC	0.691	484.1 Da LogP -2.50 TPSA 264.4	2 viol.	✓ Clean	`O=c1ccn([C@@H]2O[C@@H](CO[P@](=O)(O)O[P@](=O)(O…`
ZINC13548378 ZINC	0.691	484.1 Da LogP -2.50 TPSA 264.4	2 viol.	✓ Clean	`O=c1ccn([C@@H]2O[C@H](CO[P@@](=O)(O)O[P@@](=O)(…`
ZINC25726233 ZINC	0.691	484.1 Da LogP -2.50 TPSA 264.4	2 viol.	✓ Clean	`O=c1ccn([C@H]2O[C@@H](CO[P@](=O)(O)O[P@](=O)(O)…`
ZINC33979244 ZINC	0.691	402.2 Da LogP -1.28 TPSA 197.6	✓ Ro5	✓ Clean	`Cc1cn([C@H]2C[C@H](O)[C@H](CO[P@@](=O)(O)OP(=O)…`
ZINC33979245 ZINC	0.691	402.2 Da LogP -1.28 TPSA 197.6	✓ Ro5	✓ Clean	`Cc1cn([C@@H]2C[C@@H](O)[C@H](CO[P@@](=O)(O)OP(=…`
ZINC33979246 ZINC	0.691	402.2 Da LogP -1.28 TPSA 197.6	✓ Ro5	✓ Clean	`Cc1cn([C@H]2C[C@@H](O)[C@H](CO[P@@](=O)(O)OP(=O…`
ZINC3861755 ZINC	0.691	484.1 Da LogP -2.50 TPSA 264.4	2 viol.	✓ Clean	`O=c1ccn([C@@H]2O[C@H](CO[P@@](=O)(O)O[P@@](=O)(…`
ZINC3875255 ZINC	0.691	484.1 Da LogP -2.50 TPSA 264.4	2 viol.	✓ Clean	`O=c1ccn([C@H]2O[C@@H](CO[P@@](=O)(O)O[P@@](=O)(…`
ZINC3875256 ZINC	0.691	484.1 Da LogP -2.50 TPSA 264.4	2 viol.	✓ Clean	`O=c1ccn([C@@H]2O[C@@H](CO[P@@](=O)(O)O[P@@](=O)…`
ZINC3875257 ZINC	0.691	484.1 Da LogP -2.50 TPSA 264.4	2 viol.	✓ Clean	`O=c1ccn([C@H]2O[C@@H](CO[P@@](=O)(O)O[P@@](=O)(…`
ZINC3875258 ZINC	0.691	484.1 Da LogP -2.50 TPSA 264.4	2 viol.	✓ Clean	`O=c1ccn([C@@H]2O[C@@H](CO[P@@](=O)(O)O[P@@](=O)…`
ZINC8215882 ZINC	0.691	402.2 Da LogP -1.28 TPSA 197.6	✓ Ro5	✓ Clean	`Cc1cn([C@H]2C[C@H](O)[C@@H](CO[P@@](=O)(O)OP(=O…`
ZINC88466482 ZINC	0.691	484.1 Da LogP -2.50 TPSA 264.4	2 viol.	✓ Clean	`O=c1ccn([C@@H]2O[C@H](CO[P@@](=O)(O)O[P@@](=O)(…`
ZINC12501520 ZINC	0.688	458.5 Da LogP -0.88 TPSA 123.5	1 viol.	✓ Clean	`OCCOCCOCCOCCOCCOCCOCCOCCOCCOCCO`
ZINC3874716 ZINC	0.688	414.5 Da LogP -0.90 TPSA 114.3	✓ Ro5	✓ Clean	`OCCOCCOCCOCCOCCOCCOCCOCCOCCO`
ZINC4283769 ZINC	0.688	238.3 Da LogP -0.96 TPSA 77.4	✓ Ro5	✓ Clean	`OCCOCCOCCOCCOCCO`
ZINC4521548 ZINC	0.688	282.3 Da LogP -0.95 TPSA 86.6	✓ Ro5	✓ Clean	`OCCOCCOCCOCCOCCOCCO`
ZINC5178829 ZINC	0.688	326.4 Da LogP -0.93 TPSA 95.8	✓ Ro5	✓ Clean	`OCCOCCOCCOCCOCCOCCOCCO`
ZINC5178830 ZINC	0.688	370.4 Da LogP -0.91 TPSA 105.1	✓ Ro5	✓ Clean	`OCCOCCOCCOCCOCCOCCOCCOCCO`
ZINC5859031 ZINC	0.688	294.4 Da LogP 1.13 TPSA 55.4	✓ Ro5	✓ Clean	`CCOCCOCCOCCOCCOCCOCC`
ZINC13546305 ZINC	0.679	452.1 Da LogP -0.44 TPSA 223.9	✓ Ro5	✓ Clean	`O=c1ccn([C@@H]2CC[C@H](CO[P@](=O)(O)O[P@@](=O)(…`
ZINC13507072 ZINC	0.667	482.2 Da LogP -1.16 TPSA 244.1	2 viol.	✓ Clean	`Cc1cn([C@@H]2C[C@@H](O)[C@H](CO[P@@](=O)(O)O[P@…`
ZINC33979251 ZINC	0.667	482.2 Da LogP -1.16 TPSA 244.1	2 viol.	✓ Clean	`Cc1cn([C@@H]2C[C@H](O)[C@H](CO[P@@](=O)(O)O[P@@…`

PDB and ChEMBL records on this protein are shown in full. ChEMBL records from similar proteins are capped at the top 100 per protein (by pchembl) and ZINC at the top 50 (Tanimoto ≥ 0.5). ADME columns are descriptor-based screening flags, not experimental toxicity results.