MPC

Benchmark datasets of molecular property cliff (MPC) in ACANet paper

---

Overview of the MPC benchmark datasets

1. The 9 datasets of low sample size and narrow scaffolds (LSSNS) for molecular activity prediction LSSNS

Idx	dataset	target	target_type	size	reference
1	usp7	Ubiquitin carboxyl-terminal hydrolase 7	Protease	45	CHEMBL4251701
2	rip2	Serine/threonine-protein kinase RIPK2	Kinase	46	CHEMBL4266012; CHEMBL4130524
3	pkci	Protein kinase C iota	Kinase	48	CHEMBL4184321
4	phgdh	D-3-phosphoglycerate dehydrogenase	Other Enzyme	51	CHEMBL4373702
5	plk1	Serine/threonine-protein kinase PLK1	Kinase	73	CHEMBL4406868; CHEMBL4138231
6	ido1	Indoleamine 2,3-dioxygenase	Other Enzyme	78	CHEMBL4364294
7	rxfp1	Relaxin receptor 1	GPCR	117	CHEMBL3714716
8	braf	Serine/threonine-protein kinase B-raf	Kinase	128	CHEMBL3638563
9	mglur2	Metabotropic glutamate receptor 2	GPCR	244	CHEMBL3886984

2.The 30 datasets of high sample size and mixed scaffolds (HSSMS) for molecular activity prediction. Datasets are the molecular activity prediction benchmark datasets that from MoleculeACE

Idx	Dataset	Code	Target	Type	Target_type	Compounds	Cliffs
1	CHEMBL4792_Ki	OX2R	Orexin receptor 2	Ki	GPCR	1471	763
2	CHEMBL4616_EC50	GHSR	Ghrelin receptor	EC50	GPCR	682	330
3	CHEMBL244_Ki	FX	Coagulation factor X	Ki	Protease	3097	1350
4	CHEMBL237_EC50	KOR	Kappa opioid receptor	EC50	GPCR	955	400
5	CHEMBL3979_EC50	PPARd	Peroxisome proliferator-activated receptor delta	EC50	NR	1125	467
6	CHEMBL239_EC50	PPARa	Peroxisome proliferator-activated receptor alpha	EC50	NR	1721	709
7	CHEMBL234_Ki	D3R	Dopamine D3 receptor	Ki	GPCR	3657	1441
8	CHEMBL2047_EC50	FXR	Farnesoid X receptor	EC50	NR	631	245
9	CHEMBL219_Ki	D4R	Dopamine D4 receptor	Ki	GPCR	1859	715
10	CHEMBL264_Ki	HRH3	Histamine H3 receptor	Ki	GPCR	2862	1084
11	CHEMBL235_EC50	PPARy	Peroxisome proliferator-activated receptor gamma	EC50	NR	2349	881
12	CHEMBL236_Ki	DOR	Delta opioid receptor	Ki	GPCR	2598	965
13	CHEMBL4005_Ki	PIK3CA	PI3-kinase p110-alpha subunit	Ki	Transferase	960	351
14	CHEMBL218_EC50	CB1	Cannabinoid receptor 1	EC50	GPCR	1031	367
15	CHEMBL237_Ki	KOR	Kappa opioid receptor	Ki	GPCR	2602	941
16	CHEMBL204_Ki	F2	Thrombin	Ki	Protease	2754	989
17	CHEMBL214_Ki	5-HT1A	Serotonin 1a receptor	Ki	GPCR	3317	1147
18	CHEMBL228_Ki	SERT	Serotonin transporter	Ki	Other	1704	599
19	CHEMBL287_Ki	SOR	Sigma opioid receptor	Ki	Other	1328	464
20	CHEMBL233_Ki	MOR	u-opioid receptor	Ki	GPCR	3142	1111
21	CHEMBL2147_Ki	PIM1	Serine/threonine-protein kinase PIM1	Ki	Kinase	1456	485
22	CHEMBL1862_Ki	ABL1	Tyrosine-protein kinase ABL1	Ki	Kinase	794	253
23	CHEMBL2034_Ki	GR	Glucocorticoid receptor	Ki	NR	750	230
24	CHEMBL238_Ki	DAT	Dopamine transporter	Ki	Other	1052	263
25	CHEMBL1871_Ki	AR	Androgen Receptor	Ki	NR	659	157
26	CHEMBL231_Ki	HRH1	Histamine H1 receptor	Ki	GPCR	973	224
27	CHEMBL262_Ki	GSK3	Glycogen synthase kinase-3 beta	Ki	Kinase	856	158
28	CHEMBL2971_Ki	JAK2	Janus kinase 2	Ki	Kinase	976	120
29	CHEMBL4203_Ki	CLK4	Dual specificity protein kinase CLK4	Ki	Kinase	731	64
30	CHEMBL2835_Ki	JAK1	Janus kinase 1	Ki	Kinase	615	46

3. The 3 matched molecular pair (MMP) datasets of activity cliff classification. Datasets are from ACGCN

Dataset	Total Compounds	Total MMPs	MMP-Cliffs	MMP-nonCliffs
Thrombin	3171	5751	317	4408
Mu opioid receptor	3625	8725	219	7097
Melanocortin receptor 4	1858	7169	111	5750

4. The 10 datasets of ADMET properties in delta prediction. Datasets are from DeepDelta

Dataset	Property	Size	Units
Free Solvation	Absorption	642	Experimental Hydration Free Energy in Water
Caco2	Absorption	910	Log(Papp)
Aqueous Solubility	Absorption	1128	LogS
Fraction Unbound, Brain	Distribution	253	Log(fu,brain)
Volume of Distribution at Steady State	Distribution	1130	Log(Body/Blood Concentration in L/kg)
Renal Clearance	Excretion	636	Log(CLr)
Microsomal Clearance	Metabolism	731	Log(mL/min/kg cleared)
Hepatic Clearance	Metabolism	881	Log(mL/min/kg cleared)
Half-Life	Metabolism	1321	Log(Half-Life in Hours)
Hemolytic Toxicity	Toxicity	828	Log(HD50)