Supporting Information s17

Supporting Information

Combining Molecular Dynamics Simulation And Ligand-Receptor Contacts

Analysis As A New Approach For Pharmcophore Modeling: Beta-Secretase 1

And Check Point Kinase 1 As Case Studies

Ma'mon M. Hatmala, Shadi Jaberb, Mutasem O. Tahac*

aDepartment of Medical Laboratory Sciences, Faculty of Allied Health Sciences, The

Hashemite University, Zarqa, Jordan

bDepartment of Medical Laboratory Sciences, Faculty of Allied Health Sciences, Zar-

qa Private University, Zarqa, Jordan

cDrug Discovery Unit, Department of Pharmaceutical Sciences, Faculty of Pharmacy,

University of Jordan, Amman, Jordan

*Corresponding Author,

Telephone: 00962 6 5355000 ext. 23305.

Fax: 00962 6 5339649.

Email: [email protected]

Figure SM1: Two views of three-dimensional plot showing three main principal com- ponents calculated for Chk1 testing set (based on 12 physicochemical descriptors, see text). Red triangles (▲) represent active member compounds (IC50 ≤ 10) while blue spheres (●) represent inactive compounds (IC50 > 500) as enlisted in CHEMBL data- base.

Figure SM2: Two views of three-dimensional plot showing three main principal com- ponents calculated for BACE1 testing set (based on 12 physicochemical descriptors, see text). Red triangles (▲) represent active member compounds (IC50 ≤ 10) while blue spheres (●) represent inactive compounds (IC50 > 500) as enlisted in CHEMBL data- base. Section SM-1: ROC analysis

ROC curve is plotted by considering the highest score of an active molecule as the

first threshold then counting the number of inactive compounds within this cut-off

value, and both the corresponding sensitivity (Se) and specificity (Sp) are calculated

using equation 1 and equation 2, respectively. This process is repeated using the ac-

tive molecule possessing the second highest score and so on, until the scores of all ac-

tive compounds are considered as selection cut-off values [41].

Number of Selected Actives TP Se   ………………….. (1) Total Number of Actives TP  FN

….…….……….. (2) Number of Discarded Inactives TN Sp   Total Number of Inactives TN  FP where, TP (true positive) is the number of active compounds that are captured by the

pharmacophore under concern, FN (false negative) is the number of active com-

pounds discarded from the hits list, TN (true negative) is the number of discarded de-

coys , while FP (false positive) is the number of captured decoys .

The ROC curve for ideal distributions, where active compounds are completely sepa-

rate and distinct from the inactives (i.e., no overlap between actives and decoys), the

curve arise vertically to the upper-left corner (Se = Sp = 1) and then joins the up-

per-right corner horizontally. Hence, the more a ROC curve bends towards the upper

left corner of the diagram, the more distinct the signal appears [41].

In practice, the ROC curve for a set of actives and decoys with randomly distributed

scores tends towards the Se = 1-Sp line asymptotically with increasing number of ac-

tives and decoys.33 The success of particular virtual screening workflow depending

on ROC analysis evaluation can be provided as follow: 1) Area under the ROC curve (AUC): which range from 0.5 to 1, where 1 indicates an optimal value and 0.5 random distribution. Whereas an AUC value lower than 0.5 represents the unfavourable case of a virtual screening method that has a higher prob- ability to assign the best scores to decoys than to actives [41].

2) Overall specificity (SPC): that describes the percentage of discarded inactive by the particular virtual screening workflow. Inactive test compounds are assigned a binary score value of zero (compound not captured) or one (compound captured) [41-43].

3) Overall true positive rate (TPR or overall sensitivity): describes the fraction percentage of captured actives from the total number of actives. Active test compounds are assigned a binary score value of zero (compound not captured) or one (compound captured) [41-43].