ADMET

The previously prepared 1,008 low energy 3D chemical structures in the AfroDb library were saved in.mol2 format and initially treated with LigPrep [63] , distributed by Schrodinger Inc. This implementation was carried out with the graphical user interface (GUI) of the Maestro software package [64] , using the OPLS forcefield [65] (link)–[67] (link). Protonation states at biologically relevant pH were correctly assigned (group I metals in simple salts were disconnected, strong acids were deprotonated, strong bases protonated, while topological duplicates and explicit hydrogens were added). A set of ADMET-related properties (a total of 46 molecular descriptors) were calculated by using the QikProp program [68] running in normal mode. QikProp generates physically relevant descriptors, and uses them to perform ADMET predictions. An overall ADME-compliance score – drug-likeness parameter (indicated by #stars), was used to assess the pharmacokinetic profiles of the compounds within the AfroDb library. The #stars parameter indicates the number of property descriptors computed by QikProp that fall outside the optimum range of values for 95% of known drugs. The methods implemented were developed by Jorgensen et al. [69] (link)–[70] and among the calculated descriptors are: the total solvent-accessible molecular surface, in Ų (probe radius 1.4 Å) (range for 95% of drugs: 300–1000 Ų); the hydrophobic portion of the solvent-accessible molecular surface, in Ų (probe radius 1.4 Å) (range for 95% of drugs: 0–750 Ų); the total volume of molecule enclosed by solvent-accessible molecular surface, in ų (probe radius 1.4 Å) (range for 95% of drugs: 500–2000 ų); the logarithm of aqueous solubility, (range for 95% of drugs: −6.0 to 0.5) [69] (link), [71] (link); the logarithm of predicted binding constant to human serum albumin, (range for 95% of drugs: −1.5 to 1.2) [72] (link); the logarithm of predicted blood/brain barrier partition coefficient, logB/B (range for 95% of drugs: −3.0 to 1.0) [73] (link)–[75] (link); the predicted apparent Caco-2 cell membrane permeability (BIP_caco-2) in Boehringer–Ingelheim scale, in nm/s (range for 95% of drugs: <5 low, >100 high) [76] (link)–[78] (link); the predicted apparent Madin-Darby canine kidney (MDCK) cell permeability in nm s⁻¹ (<25 poor, >500 great) [77] (link); the index of cohesion interaction in solids, Ind_coh, calculated from the HBA, HBD and the surface area accessible to the solvent, SASA ( ) by the relation Ind_coh = HBA HBD^1/2/ (0.0 to 0.05 for 95% of drugs) [71] (link); the globularity descriptor, Glob =  , where r is the radius of the sphere whose volume is equal to the molecular volume (0.75 to 0.95 for 95% of drugs); the predicted polarizability, (13.0 to 70.0 for 95% of drugs); the predicted logarithm of IC₅₀ value for blockage of HERG K⁺ channels, logHERG (concern<−5) [79] (link)–[80] (link); the predicted skin permeability, (−8.0 to −1.0 for 95% of drugs) [81] (link)–[82] (link); and the number of likely metabolic reactions, #metab (range for 95% of drugs: 1–8).

In the data collection process, we finally obtained 31 datasets for ADMET modeling from the first part of data. For these datasets, the following pretreatments were carried out to guarantee the quality and reliability of the data: (1) removing compounds that without explicit description for ADME/T properties; (2) for the classification data, reserve only one entity if there are two or more same compounds; (3) for the regression data, if there are two or more entries for a molecule, the arithmetic mean value of these values was adopted to reduce the random error when their fluctuations was in a reasonable limit, otherwise, this compound would be deleted; (4) Washing molecules by MOE (disconnecting groups/metals in simple salts, keeping the largest molecular fragment and add explicit hydrogen). After that, a series of high-quality datasets were obtained. According to the Organization for Economic Co-operation and Development (OECD) principles, not only the internal validation is needed to verify the reliability and predictive ability of models, but also the external validation [11 (link)]. Therefore, all the datasets were divided into training set and test set according to the chemical space distribution by “Diverse training set split” module from ChemSAR [26 (link)]. In this step, we set a threshold that 75% compounds were used as training set and the remaining 25% as test set. The detailed information for these datasets can be seen in Table 1.Table 1

The statistical results of the datasets for modeling

Category	Property	Total	Positive	Negative	Train	Test
Basic physicochemical property	LogS	5220	–	–	4116	1104
	LogD7.4	1031	–	–	773	258
	LogP
Absorption	Caco-2	1182	–	–	886	296
	Pgp-inhibitor	2297	1372	925	1723	574
	Pgp-substrate	1252	643	609	939	313
	HIA	970	818	152	728	242
	F (20%)	1013	759	254	760	253
	F (30%)	1013	672	341	760	253
Distribution	PPB	1822	–	–	1368	454
	VD	544	–	–	408	136
	BBB	2237	540	1697	1678	559
Metabolism	CYP 1A2-inhibitor	12,145	5713	6432	9145	3000
	CYP 1A2-substrate	396	198	198	297	99
	CYP 3A4-inhibitor	11,893	5047	6846	8893	3000
	CYP 3A4-substrate	1020	510	510	765	255
	CYP 2C9-inhibitor	11,720	3960	7760	8720	3000
	CYP 2C9-substrate	784	278	506	626	156
	CYP 2C19-inhibitor	12,272	5670	6602	9272	3000
	CYP 2C19-substrate	312	156	156	234	78
	CYP 2D6-inhibitor	12,726	2342	10,384	9726	3000
	CYP 2D6-substrate	816	352	464	611	205
Excretion	Clearance	544	–	–	408	136
	T1/2	544	–	–	408	136
Toxicity	hERG	655	451	204	392	263
	H-HT	2171	1435	736	1628	543
	Ames	7619	4252	3367	5714	1905
	Skin sensitivity	404	274	130	323	81
	Rat oral acute toxicity	7397	–	–	5917	1480
	DILI	475	236	239	380	95
	FDAMDD	803	442	361	643	160

Left tibias were brought to room temperature before testing and kept hydrated in calcium-buffered saline until the test was complete. Bones were tested in the ML direction (medial surface in tension) in four-point bending (Admet eXpert 450 Universal Testing Machine; Norwood, MA, USA). The fibula was carefully removed from each bone using a scalpel, and the bones were positioned with the TFJ aligned with the outside edge of one loading roller, preloaded to 0.5 N, preconditioned for 15 s (2 Hz, mean load of 2 ± 2 N), and monotonically tested to failure in displacement control at a rate of 0.025 mm/s. Load and deflection were recorded, from which structural strength (yield and ultimate forces), stiffness (slope of the linear portion of the force versus displacement curve), and deformation (yield deformation, postyield deformation, and total deformation) were determined.(11 (link),31 (link))
Bones were visually monitored during testing, and the point of fracture initiation was measured relative to the proximal end. A subset of geometric properties at the fracture site was obtained from μCT data (I_AP and the distance from the centroid to the tensile surface of the bone, c). Together with the load and deflection data, I_AP and c were used to map force and displacement (structural-level properties dependent on bone structural organization) into stress and strain (predicted tissue-level properties) from standard beam-bending equations for four-point bending:

In these equations, F is the force, d is the displacement, a is the distance from the support to the inner loading point (3 mm), and L is the span between the outer supports (9 mm). The yield point was calculated using the 0.2% offset method based on the stress-strain curve. The modulus of elasticity was calculated as the slope of the linear portion of the stress-strain curve.

The drug-likeness analysis was performed by admetSAR and SwissADME to confirm any cytotoxicity produced by ligands in humans. Several pharmacokinetics properties such as absorption, distribution, metabolism, excretion and toxicity (ADMET) of the tested compound chrysin and the positive control lamivudine were measured with the web tools admetSAR [37 (link)] and SwissADME [38 (link)]. The physicochemical properties were also studied with the help of these servers.

The ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) approach was used for in-silico pharmacokinetic predictions of the selected PTB drugs. ADMET lab platform (http://admet.scbdd.com/home/index/) [56 (link)] was used to access the ADMET properties. The assessment was carried out for each physiochemical property by submitting a SMILE format of the individual compound.