3582279669

Wirtualne przesiewanie: brak wiedzy o strukturze trójwymiarowej celu terapeutycznego.

Tubie 4. RkiD and Prwldon Avm|fd(htr Dalawtt for Flvv larjru 1120]

	S\3I	RF	KNN	DT	ASS	TV	NB
		Diliul 1: Ohw		i'ed A<ll<n( ' - 4.2H
S Kwil	*2(4)	U(«)	*1(4)	•»(7)	71(14)	*5(21)	50 (10)
H hwnm	•6(13)	#7(4)	4* (15)	41 C*>	60(10)	•7(7)	)2 (2))
			Diiiwt II: Obwntd Aidin a.l -		MH
S RecaU	*9(18)	45(23)	*1 ()0)	45 00)	15(15)	3* (25)	25(40)
H Prrcnioo	W (14)	*5(22)	)l (10)	2J(I2)	12(15)	55 (2*)	12(4)

TablrS. Rwali and PrwMon Avera|«d (her Dalatrl* for Four TargH» |I20|

	Krcafl		PiwHIoa
	1/3	2/3	vy	2/3
SVM	48 2 (12 9)	5*4(27 2)	52 3(16 2)	36 2(14 1)
RF	15.4(10 0)	393 (14 9)	81 9(24 5)	79 8(23 5)

Tbr Uaa| wt cootunrd nta Ar car ttod (13) oc Ar Twe &sdt (2/)) of com-pouodt paarokd fuvl Stindud drwuboat o biłckrtł

[DMP4] “Performance of machinę learning methods for ligand-based virtual screening" by D. Plewczynski, S. Spieser and U. Koch. Combinatorial Chemistry & High Throughput Screening CCHTS 12(4): 358-68 (2009);

[DMP5] “Assessing Different Classification Methods for Virtual Screening" by D. Plewczynski, S. Spieser, U. Koch. J. Chem. Inf. Model. 46(3), p.1098-106 (2006);