Informatics in Medicine Unlocked 37 (2023) 101162
Contents lists available at ScienceDirect 
Informatics in Medicine Unlocked 
journal homepage: www.elsevier.com/locate/imu 
Targeting Leishmania donovani sterol methyltransferase for leads using 
pharmacophore modeling and computational molecular mechanics studies 
Patrick O. Sakyi a,b, Emmanuel Broni c,d,e, Richard K. Amewu a,**, Whelton A. Miller III e,f,g, 
Michael D. Wilson d,e, Samuel K. Kwofie c,h,* 
a Department of Chemistry, School of Physical and Mathematical Sciences, College of Basic and Applied Sciences, University of Ghana, P. O. BOX LG 56, Legon, Accra, 
Ghana 
b Department of Chemical Sciences, School of Sciences, University of Energy and Natural Resources, Box 214, Sunyani, Ghana 
c Department of Biomedical Engineering, School of Engineering Sciences, College of Basic & Applied Sciences, University of Ghana, Legon, Accra, LG 77, Ghana 
d Department of Parasitology, Noguchi Memorial Institute for Medical Research (NMIMR), College of Health Sciences (CHS), University of Ghana, Legon, Accra, LG 581, 
Ghana 
e Department of Medicine, Loyola University Medical Center, Maywood, IL, 60153, USA 
f Department of Molecular Pharmacology and Neuroscience, Loyola University Medical Center, Maywood, IL, 60153, USA 
g Department of Chemical and Biomolecular Engineering, School of Engineering and Applied Science, University of Pennsylvania, Philadelphia, PA, 19104, USA 
h Department of Biochemistry, Cell and Molecular Biology, West African Centre for Cell Biology of Infectious Pathogens, College of Basic and Applied Sciences, University 
of Ghana, P.O. Box LG 54, Accra, Ghana   
A R T I C L E  I N F O   A B S T R A C T   
Keywords: The mortalities and morbidities of leishmaniasis are high and the disease is under reported globally. The absence of 
Leishmania donovani vaccines coupled with chemotherapeutic challenges including chemoresistance, scarcity and toxicity have made the 
Sterol methyltransferase fight against leishmaniasis an arduous one. Furthermore, the treatment options currently available for leishmaniasis 
Pharmacophore are long and sometimes require hospitalization. There is therefore the need to explore novel pathways to identify new 
Ergosterol biosynthesis 
Molecular docking compounds with alternative mechanisms of action. A pharmacophore-based screening was employed in identifying 
Molecular mechanics new potential inhibitors with unique scaffolds targeting Leishmania donovani sterol methyltransferase (LdSMT), a key 
enzyme for ergosterol biosynthesis. To accomplish this, 22,26-azasterol, a known inhibitor of this target and five 
other derivatives with IC50 less than 10 μM were used to generate a robust 3D pharmacophore model via LigandScout 
with a score of 0.9144. The validated model was used as a query to screen a library of 69034 natural products 
obtained from the InterBioScreen Limited. Compounds with pharmacophore fit scores above 50 were docked against 
the modelled structure of LdSMT. Altogether, ten molecules with binding energies between − 7 and − 11 kcal/mol 
were identified as potential bioactive molecules. The molecular dynamics simulation and molecular mechanics 
Poisson–Boltzmann surface area (MM-PBSA) calculations reinforced the results from the docking studies suggesting 
the selected hits bind effectively at the active sites of the target protein. The compounds were observed to bind in the 
S-adenosine-L-homocysteine binding pocket of the modelled LdSMT with Trp208 and Val330 predicted as key res-
idues critical for ligand binding. Prediction of biological activity with probability of activity (Pa) greater than 
probability of inactivity (Pi) revealed that seven compounds (STOCKIN-54848, STOCKIN-89115, STOCKIN-68720, 
STOCKIN-44724, STOCKIN-76694, STOCKIN-47277 and STOCKIN-95708) possessed antileishmanial properties. 
STOCKIN-89115, STOCKIN-68720, STOCKIN-44724, and STOCKIN-47277 were predicted to be membrane perme-
ability inhibitors, while all ten hit compounds possessed antineoplastic activity. The compounds have the propensity 
of disrupting ergosterol biosynthesis leading to the suppression of growth in Leishmania donovani. The compounds 
were predicted to have good absorption, distribution, metabolism, excretion and toxicity profiles, hence their po-
tential antileishmanial activity can be exploited upon experimental corroboration.  
* Corresponding author. Department of Biomedical Engineering, School of Engineering Sciences, College of Basic & Applied Sciences, University of Ghana, Legon, 
Accra, LG 77, Ghana. 
** Corresponding author. 
E-mail addresses: opsakyi@st.ug.edu.gh, patrick.sakyi@uenr.edu.gh (P.O. Sakyi), ebroni002@st.ug.edu.gh (E. Broni), ramewu@ug.edu.gh (R.K. Amewu), 
wmiller6@luc.edu (W.A. Miller), MWilson@noguchi.ug.edu.gh (M.D. Wilson), skkwofie@ug.edu.gh (S.K. Kwofie).  
https://doi.org/10.1016/j.imu.2023.101162 
Received 7 September 2022; Received in revised form 30 December 2022; Accepted 3 January 2023   
Available online 7 January 2023
2352-9148/© 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
P.O. Sakyi et al.                                                                                                                                                                                   I n  f o  r m   a t i c  s  i n   M   e d  i c i n  e   U  n  l o c  k  e d 37 (2023) 101162
1. Introduction There are currently no vaccines for combating VL and all other forms 
of leishmaniasis as the current drugs used to treat the disease including 
Leishmaniasis, caused by over twenty Leishmania species and trans- pentavalent antimony, pentamidine (PTM), amphotericin B (Amp B), 
mitted by more than eighty sandfly species, is second to malaria in terms miltefosine (Milt), paromomycin and liposomal Amp B are plagued with 
of mortalities among all the neglected tropical diseases (NTDs) [1]. numerous challenges such as long treatment durations, cytotoxicity, 
Visceral leishmaniasis (VL) caused mainly by L. donovani and L. infantum resistance, and high cost [19,20]. Synergistic resistance has also been 
is the most fatal among the three forms of leishmaniasis and is mostly observed in partner drugs used in combinational chemotherapies. As a 
characterized by irregular bouts of fever, weight loss, enlargement of the result there is the need to find new chemotypes by exploring novel 
liver, and anemia [1,2]. Even more disturbing is the presence of macu- pathways to identify new compounds with alternative mechanisms of 
lar, maculopapular, and nodular rash in patients who have undergone action from the existing drugs [7]. In our earlier study, we employed de 
successful leishmaniasis treatment [3]. Most of the recovered patients novo design to predict inhibitors against SMT [21]. 
suffer from inferiority complex due to the complications hence struggle Pharmacophore-based drug design is an alternative to de novo design. 
to integrate into society because of the ugly scars [4]. This approach though cheaper, has however been employed in rational 
VL is endemic in poverty-stricken parts of the world and an estimated drug design to overcome most of the challenges associated with drug 
50,000 to 90,000 new cases occur worldwide annually [5]. Most of the inefficiencies and resistance [22]. In addition, pharmacophore-based 
cases are recorded in Brazil, Ethiopia, Eritrea, India, Iraq, Kenya, Nepal, drug design explores diversity thus increases the chance of attaining 
Somalia, South Sudan and Sudan [6,7]. Socioeconomic conditions, leads and promotes specificity [23,24] in the design of potent inhibitors 
malnutrition, population mobility, environmental and climate change against plausible targets. Pharmacophore, an ensemble of steric and 
are the major risk factors associated with this disease [8]. Leishmania electronic features in a molecule recognizable at a protein’s binding site 
parasites are transmitted through the bites of infected female phlebot- and responsible for a molecule’s biological activity [22], has been 
omine sandflies, which feed on blood to produce eggs [9]. The epide- employed in recent strategies for the design of drugs for treatment of 
miology of VL is complex and depends on parasite characteristics and cancer (gefitinib), inflammatory diseases (morazone), microbial in-
sandfly species. Local ecological characteristics of the transmission sites, fections (finafloxacin) and psychotic diseases (phendimetrazine) [25]. 
current and past exposure of the human population and behaviors have Interestingly, most of the pharmacophore-based drugs exhibit multi-
contributed to the challenges associated with VL treatments. Moreover, modality properties which gives an added advantage and hence helps to 
lack of comprehensive understanding of the ergosterol, the main sterol overcome the parasite’s resistance [26]. Though yet to be fully explored 
biosynthesized in Leishmania parasites and their essentiality for the for the design of drugs for leishmaniasis treatment, irosustat, avasimibe, 
parasite’s survival have also made the treatment of VL a challenging topiramate and RWJ-37497 used for cancer treatments targeting two or 
one. more biological targets, were designed based on the core aryl O-sulfa-
Sterol methyltransferase (SMT), a potential target for the treatment mate pharmacophore [26]. Similarly, linezolid, timolol, merazone and 
of leishmaniasis, catalyzes the transfer of a methyl group from S-aden- phendimetrazine used as antibiotic, antihypertensive, 
osine-methionine to the C24 position of the sterol side chain during anti-inflammatory and antipsychotic drugs, respectively were all 
ergosterol biosynthesis (Fig. 1) [7]. This target has provided the focus designed with the morpholine pharmacophore [25]. 
for study of electrophilic alkylations, a reaction type of functional Natural products from various sources have been explored for the 
importance that results in the formation of C–C bond in natural products treatments of numerous debilitating diseases affecting humans [27]. Till 
[10]. Inhibitors of SMT have shown antiproliferative effects against date, nature has been one of the main sources of drugs and between 
pathogenic fungi [11,12], Trypanosoma cruzi [13] and T. brucei [14,15]. 1981 and 2014, about 50% of all US FDA drugs have had their source 
Moreover, 22,26-azasterol, a known SMT inhibitor induces drastic from molecules isolated from natural sources [28]. This notwith-
changes in the sterol profile of L. donovani and L. amazonensis charac- standing, their synthetic accessibility has hampered their use in some 
terized by the loss of ergostane-based sterols and their replacements by instances, but their structural diversity and low toxicity compared to 
24-desalkyl sterols, along with significant growth reduction and subse- their synthetic counterparts have made them the most preferred choice 
quently elimination of the parasite [11]. In addition, SMT inhibitors also in the search for new drugs [27]. Exploring key pharmacophores from 
cause profound structural changes in the plasma membrane, mito- natural product isolates could provide a possible means of finding 
chondrial membrane, and endoplasmic reticulum of L. amazonensis suitable druggable candidates for leishmaniasis treatment. 
[16–18]. The effects of SMT inhibition on parasite’s survival coupled Since the essentiality of SMT to Leishmania parasites is no more in 
with the absence of SMT orthologues in humans makes SMT a potential doubt and it is suggested as a plausible target, herein, we present the 
target for exploitation therapeutically for drug design against Leishmania application of ligand-based pharmacophore screening and molecular 
parasites. docking to identify potential inhibitors from natural products with 
Fig. 1. 24-SMT catalyzes the formation of ergosterol from zymosterol.  
2
P.O. Sakyi et al.                                                                                                                                                                                   I n  f o  r m   a t i c  s  i n   M   e d  i c i n  e   U  n  l o c  k  e d 37 (2023) 101162
unique scaffolds possessing the possibility of ameliorating the activity of 2.3. Compound selection for pharmacophore generation 
LdSMT. The study also seeks to predict the mechanism of binding and 
biological activity of the proposed compounds. A search through literature identified 22,26-azasterol and its de-
rivatives, imipramine and azetimibe (Table 1) as inhibitors of SMT [17, 
2. Materials and methods 53,54]. ChemDraw Ultra 12.0 [36] was used to generate the 2D of the 
chemical structures and Chem3D Pro 12.0 [36] was utilized to generate 
The scheme (Fig. 2) describes the detailed procedure adopted in the energy minimized 3D structures. The compound 22,26-azasterol, a 
prediction of potential inhibitors from natural products targeting known inhibitor of the target protein was specifically docked to the 
LdSMT. Six ligands including 22,26-azasterol with IC50 less than 10 μM model protein to determine binding mechanisms and consistency within 
were used to generate a pharmacophore model using the merged feature the active site. 
on LigandScout version 4.3 [29]. The pharmacophore model was used to 
screen a library of natural products from InterBioScreen Limited and 2.4. Ligand-based pharmacophore virtual screening 
ligands with pharmacophore fit scores greater than 50 were subjected to 
molecular docking studies using a modelled LdSMT structure as the LigandScout version 4.3 [29] was used to generate the pharmaco-
target protein [21]. The potential lead compounds were further evalu- phore model needed for the ligand-based pharmacophore virtual 
ated for their physicochemical, pharmacological and toxicity profiles, screening. The 3D structures of the SMT inhibitors with IC50 less than 10 
biological activity predictions and molecular dynamics (MD) μM were uploaded onto LigandScout’s Ligand-Based Modeling 
computations. Perspective v4.3 [29] using their Structure Data File (SDF) formats. The 
default settings of OMEGA best in LigandScout version 4.3 [29] was used 
2.1. Protein structure preparation in the generation of ligand conformations with 200 conformations being 
the maximum limit set per molecule. 
The modelled 3D structure of LdSMT using Modeller version 10.2 
[30] in our previous study [21] was used for this study. The preparation 2.5. Library retrieval and preparation for pharmacophore-based screening 
of the protein structure has been previously described [21] and pre-
sented here. The 3D LdSMT structure was energy minimized using the A natural product chemical library from InterBioScreen comprised of 
Optimized Potentials for Liquid Simulations (OPLS)/All Atom (AA) force 69034 compounds was retrieved [37]. The chemical entities were used 
filed in GROMACS 2018 [31,32]. Biovia Discovery Studio Visualizer for the pharmacophore-based virtual screening. 
v19.1.0.18287 [33] was used to visualize the energy minimized struc-
ture as well as remove water molecules solvating the protein. The 
structure was then saved in Protein Data Bank (pdb) format using the 2.6. Pharmacophore-based virtual screening of the libraries 
Biovia Discovery Studio Visualizer v19.1.0.18287 [33]. The structure 
was then converted to AutoDock Vina’s [34] compatible “Protein Data In all, 69034 chemical entities were used for the pharmacophore 
Bank, Partial Charge (Q), & Atom Type (T) (pdbqt)” format for molec- screening using LigandScout v.4.3 [29] by first converting from “sdf” to 
ular docking. “lbd” before screening against a validated pharmacophore model. 
2.2. Prediction of binding sites 2.7. Validation of both pharmacophore model and AutoDock Vina 
The probable volume and area of the binding site of the modelled Both the pharmacophore model and docking protocol were validated 
protein was determined using the Computed Atlas of Surface Topog- before the virtual screening. 
raphy of proteins (CASTp) [35], and was described previously [21] and 
described here. Amino acid residues present in the binding sites were 2.7.1. Validation of pharmacophore model 
also identified and visualized using Biovia Discovery Studio Visualizer Validation of the pharmacophore model in LigandScout v.4.3 [29] 
v19.1.0.18287 [33]. was done using the receiver operating characteristic (ROC) curve and 
the Enrichment Factor (EF). This was undertaken to ascertain the ability 
of the pharmacophore model to distinguish between actives and decoys. 
To do this, the six inhibitors with IC50 < 10 μM (Table 1) against SMT 
Fig. 2. The workflow employed in predicting potential antileishmanial compounds targeting LdSMT. The scheme provides a pharmacophore-based design coupled 
with molecular docking and dynamics simulations to identify natural product chemotypes with potential leishmanicidal properties. 
3
P.O. Sakyi et al.                                                                                                                                                                                   I n  f o  r m   a t i c  s  i n   M   e d  i c i n  e   U  n  l o c  k  e d 37 (2023) 101162
Table 1 2.8. Molecular docking studies of chemical entities with good 
Chemical structures of six known inhibitors of Leishmania parasite sterol meth- pharmacophore fit scores 
yltransferase and their IC50 (μM).  
Compound IC50/μM Reference AutoDock Vina [34] interfaced with PyRx v.0.8 [41] was used for the 
8.90 [14] virtual screening as described previously [21]. Ligands with good 
pharmacophore fit scores obtained after screening the pharmacophore 
model against the library were imported as spatial data file (sdf) format 
into AutoDock Vina, energy minimized and converted to pdbqt format. 
The energy minimization of ligands employed default settings involving 
universal force field (UFF) and conjugate gradient for optimization al-
gorithm for 200 steps. The energy minimized ligands were then docked 
4.80 [14] against the energy minimized LdSMT using AutoDock Vina [34]. The 
grid box size was set as previous work [21] to 91.445 × 73.502 ×
78.352 Å3 with the center at (72.200, 58.009, 13.302) Å. Ligands were 
then docked into the binding site of the target protein with exhaus-
tiveness set to 8. 
2.9. Characterization of mechanism of binding 
2.50 [14] 
The atomistic details of binding between LdSMT and the ligands were 
determined using the Biovia Discovery Studio Visualizer v19.1.0.18287 
[33] as described previously [21]. 
2.10. ADMET properties and drug-likeness assessment 
3.85 [14] 
The absorption, distribution, metabolism, excretion, and toxicity 
(ADMET) properties of shortlisted compounds were determined using 
SwissADME [42] and OSIRIS Property Explorer in Data Warrior 5.0.0 
[43] for the toxicity effects (mutagenic, tumorigenic, irritant, and 
reproductive). PAINS and synthetic accessibility were further performed 
to eliminate false positive compounds that possess good physicochem-
ical properties as well as compounds difficult to synthesize [42]. 
7.70 [14] 2.11. Prediction of biological activity of selected compounds 
The biological activity of the selected compounds were predicted 
using the prediction of activity spectra for substances (PASS) [44], 
which works based on a training dataset of known substances present in 
its database. The simplified molecular input line entry system (SMILES) 
8.90 [14]  files of shortlisted molecules were used as inputs. 
2.12. Molecular dynamics simulation and MM-PBSA evaluation of 
LdSMT-Ligand complexes 
The MD simulation and molecular mechanics Poisson–Boltzmann 
surface area (MM-PBSA) were computed as previously described [21] 
using GROMACS 2018 [31,32] and g_mmpbsa [45], respectively. The 
and their decoys were used to generate both the ROC curve and EF. The root mean square deviation (RMSD), root mean square fluctuation 
ROC curve is quantified by the calculation of area under the curve (AUC) (RMSF), radius of gyration (Rg) and solvent accessible surface area 
with values ranging from 0 to 1 and the EF measures the ratio of actives (SASA) graphs [46] were plotted using Qtgrace [47], while the 
in the top hits of a virtual screen to that expected by random selection, per-residue energy decomposition graphs were generated using R pro-
expressed as a percentage [38]. gramming language [48,49]. 
2.7.2. Validation of AutoDock Vina 3. Results and discussion 
The six known inhibitors of SMT comprising 22,26-azasterol and its 
derivatives (Table 1) served as actives for SMT. The SMILES of the six The results of the pharmacophore modelling, molecular docking, 
compounds served as inputs for the generation of 50 decoys each via ADMET evaluation, prediction of biological activities of the selected 
Directory of Useful Decoys (DUD-E) [39]. The 306 compounds molecules, MD simulations and MM-PBSA calculations are presented. 
comprising 6 actives and 300 decoys were docked to the catalytic 
domain of the LdSMT. EasyROC version 1.3.3 [40] was used to compute 3.1. Prediction of LdSMT active site residues 
the AUC from the ROC curve employing the respective binding energies 
of the ligands. The calculated AUC was used to validate the ability of Proteins are important class of biological macromolecules which 
AutoDock Vina to differentiate actives from decoys. realize their functions by binding to other proteins or small molecules 
[50]. Understanding protein-ligand interactions is therefore, very 
essential for studying the biology of the protein at the molecular level 
and for designing more selective and potent inhibitors against target 
4
P.O. Sakyi et al.                                                                                                                                                                                   I n  f o  r m   a t i c  s  i n   M   e d  i c i n  e   U  n  l o c  k  e d 37 (2023) 101162
proteins [50]. To achieve a better understanding, the binding pocket of a used for the model generation. The merge-feature model setting was 
receptor for ligand binding is very crucial. As such, the volume and area used in this research instead of the shared-feature because the 
of the active site in LdSMT was computed via Computed Atlas of Surface merge-feature selects all the chemical descriptors present on each ligand 
Topography of proteins (CASTp) [35]. A binding site volume of 446.632 even when the descriptors are not common to most of the ligands in the 
Å3 and area of 905.262 Å2 were obtained from previous studies [21] training set [29]. The pharmacophore model generated is scored from 
with residue positions spanning Gly4, Gly5, Arg6, Glu7, Thr8, Ala30, 0 to 1 with scores very close to 1 being the most accurate model to 
Arg34, Thr51, Met52, Val53, Asn54, Tyr66, Gly67, Asp108, Val109, distinguish actives from inactives. The generated pharmacophore model 
Gly110, Cys111, Gly112, Val113, Gly114, Gly115, Pro116, Ala117, was predicted to have a score of 0.9144 based on the merged features 
Asn132, Gln137, Tyr175, Ala176, Ile177, Glu178, Ala179, Thr180, from all six ligands and with the score close to 1, it suggests the model 
Cys181, His182, Lys186, Cys189, Tyr190, Val193, Phe203, Leu205, can accurately predict actives from the chemical library. The features 
Tyr206, Gly207, Trp208, Met210, Tyr214, Pro216, Asn217, Asp218, generated from the pharmacophore modeling consisted of three hy-
Glu219, Tyr220, Arg222, Ile224, Lys225, His226,Arg227, Ile228, drophobic interactions, one positive ionizable, one hydrogen bond 
Glu229, Leu230, Glu236, Lys241, Met247, Phe252, Ile261, Il269, acceptor and one hydrogen bond donor (Fig. 3). 
Ser271, Ile272, Trp274, Tyr275, Leu278, Glu324, Ser328, Leu329, 
Val330, Val331, Gly332, Gly333 and Leu335 [21]. 3.3. Validation of pharmacophore model and docking protocol 
3.2. Pharmacophore generation The ROC curve was employed to validate the pharmacophore model 
and docking protocol. Six inhibitors of the SMT were used to generate 
Six of the ligands (Table 1) with IC50 less than 10 μM were used to 300 decoys from Database of Useful Decoys: Enhanced (DUD-E) [39], 
generate the pharmacophore model. Previous studies have used mainly with 50 decoy generated from each ligand. A library comprising six 
two ligands in generating the pharmacophore model [38,46] but to actives and the decoys labelled as “actives” and “inactives”, respectively 
achieve a significantly robust pharmacophore model, 48 inhibitors were in LigandScout were screened using the best generated pharmacophore 
used for generating the model for the design of inhibitors against histone model with score 0.9144. 
deacetylase 2 (HDAC2) [51]. Similarly, to ensure statistical relevance, a 
training set containing 18 diverse compounds with experimental activity 3.3.1. Validation of generated pharmacophore model 
values (IC50) ranging from 0.5 nM to 5590 nM were employed in the A good pharmacophore model should be able to differentiate be-
development, evaluation and application of 3D quantitative structural tween actives and inactives (the decoys) from a library of compounds 
activity relationship (QSAR) pharmacophore model in the discovery of [38]. The ROC measures the performance of the pharmacophore model 
potential human renin inhibitors [52]. In the current study, six com- to effectively differentiate actives from inactives. AUC is assigned values 
pounds (22,26-azasterol and its derivatives) with IC50 values less than from 0 to 1 with values very close to 1 signifying the model can effec-
10 μM were used as the training set for the generation of a robust and a tively select actives from a library and 0 meaning incorrect classification 
statistically relevant pharmacophore model. The inherent features of the [38]. AUC of 0.5 implies poor classification between actives and in-
compounds were identified to generate the hypothetical model. actives from a chemical library [38]. The AUC was determined as 1.0, 
LigandScout [29] allows the generation of 3D pharmacophore model 1.0, 1.0, and 1.0 in the top 1%, 5%, 10% and 100% (Fig. 4) of the 
from either the 2D or 3D structural data for any number of ligands screened library, respectively. Altogether, the AUC was observed to be 
compared to other programs which restrict the number of ligands to be 1.0 implying that the pharmacophore model can efficiently classify 
Fig. 3. The pharmacophore features generated from (a) the six ligands comprising hydrophobic interactions (H), positive ionizable (PI), hydrogen bond acceptor 
(HBA) and hydrogen bond donor (HBD), (b) the pharmacophore model produced from merging the pharmacophore descriptors of all six inhibitors. Red balls showing 
hydrogen bond acceptors (HBA), yellow balls representing hydrophobic interactions (H), green regions showing hydrogen bond donors (HBD) and blue regions 
representing positive ionizable (PI), and (c) represents the superimposition of the sixteen selected hits on the generated pharmacophore model showing the various 
regions they align to the model. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) 
5
P.O. Sakyi et al.                                                                                                                                                                                   I n  f o  r m   a t i c  s  i n   M   e d  i c i n  e   U  n  l o c  k  e d 37 (2023) 101162
actives from the chemical library. The enrichment factors (EFs) were 
determined as 52.0, 26.0, 26.0 and 26.0 for 1%, 5%, 10% and 100% 
(Fig. 4), respectively. It supports the efficient classification ability of the 
model. 
3.3.2. Validation of docking protocol 
The ROC curve was also used to validate the docking protocol of 
AutoDock Vina [34] to ensure its ability to discriminate actives from 
inactives. The computed AUC under the ROC curve ranging from 0 to 1 
evaluates the performance of AutoDock Vina in classifying actives from 
decoys. AUCs close to 1 are considered as perfect and below 0.5 as poor 
[55,56]. The 6 inhibitors of SMT and their decoys were screened against 
the LdSMT using AutoDock Vina to generate the ROC curve. The 
computed AUC under the ROC curve was obtained to be 0.9997 (Fig. 5), 
classified as perfect [56,57], and this suggests AutoDock Vina reason-
ably differentiates actives from inactive compounds of LdSMT. A recent 
in vitro study corroborates this result as AutoDock Vina was employed Fig. 5. ROC generated after screening 6 actives and 300 decoys against LdSMT 
successfully to screen FDA drugs against LdSMT and simeprevir was producing an AUC of 0.9997. 
identified to inhibit LdSMT with an IC50 of 51.49 μM [58]. 
obtained from the pharmacophore screening and the known inhibitors, 
3.3.3. Pharmacophore-based screening of library X2 and 22,26-azasterol were docked to the binding site of the LdSMT. 
The validated pharmacophore model was used as a 3D query to The binding energies obtained for the LdSMT-22,26-azasterol and 
screen the chemical library of 69034 natural product compounds. LdSMT-X2 complexes were − 7.6 and − 7.0 kcal/mol, respectively, 
Compounds were matched and filtered based on the pharmacophore fit similar to previous studies [21]. A potential hit compound should have a 
score of 50 and above for further analysis as done previously [38]. binding energy comparable to or lower than that of known inhibitors. 
Compounds with steroidal core or synthetic accessibility more than 6 Furthermore, the potential hit compound should bind at the active site 
were also withdrawn from further studies. Steroidal core compounds of the target protein. Similar filtration criteria have already been used in 
were removed because they have the potential of causing off-target ef- the in silico identification of potential inhibitors against plausible targets 
fects consistent with earlier studies [14,59,60]. In addition, compounds [65–67]. The 66 compounds were shortlisted to 10, based on compounds 
with synthetic accessibility greater than 6 are predicted as difficult to whose binding energies were − 7.6 kcal/mol or lower, as well as the 
synthesize [61] and hence were also withdrawn. In all, 66 compounds plausible docking of the ligands at the receptor binding site. Amongst 
(Supplementary Table 1) had pharmacophore fit scores above 50 and the top 10, STOCKIN-54848 had the least binding energy of − 10.1 
hence were selected for downstream studies. The highest and least kcal/mol, while STOCKIN-86259 and STOCKIN-95708 had the highest 
pharmacophore fit scores observed were 60.56 and 51.80 for with both having − 7.6 kcal/mol. This suggests the compounds may be 
STOCKIN-63567 (Table 2) and STOCKIN-57722, respectively (Supple- good LdSMT binders and may also have the potential of inhibiting 
mentary Table 1). LdSMT. STOCKIN-63567 with the highest pharmacophore fit score of 
60.56 was eliminated from the downstream analysis since it had a high 
3.4. Molecular docking of hits from pharmacophore screening binding energy of − 6.3 kcal/mol indicating a lower binding affinity. 
The binding energy associated with the docked protein-ligand com-
plex is used to identify potential lead compounds against the target of 
interest [58,62–64]. The 66 compounds (Supplementary Table 1) 
Fig. 4. The ROC of the selected pharmacophore model. The AUCs are 1.0, 1.0, 1.0, and 1.0 in the top 1%, 5%, 10% and 100%, respectively. The EFs at 1%, 5%, 10% 
and 100% are 52.0, 26.0, 26.0 and 26.0, respectively. The further away of the median (that is, the dotted line) from the curve signifies a good model. 
6
P.O. Sakyi et al.                                                                                                                                                                                   I n  f o  r m   a t i c  s  i n   M   e d  i c i n  e   U  n  l o c  k  e d 37 (2023) 101162
Table 2 
The top 20 compounds arranged in order of decreasing pharmacophore fit scores. Downstream analysis was based on all 66 compounds.  
Compound ID Pharmacophore fit score Binding energy (kcal/mol) Compound ID Pharmacophore fit score Binding energy (kcal/mol) 
STOCKIN-63567 60.56 − 6.3 STOCKIN-89223 52.96 − 7.6 
STOCKIN-26015 60.17 − 6.1 STOCKIN-58769 52.96 − 6.8 
STOCKIN-89115 53.98 − 8.9 STOCKIN-32756 52.96 − 7.1 
STOCKIN-88290 53.47 − 7.8 STOCKIN-61176 52.90 − 8.0 
STOCKIN-90413 53.42 − 8.6 STOCKIN-92854 52.88 − 7.7 
STOCKIN-25055 53.35 − 6.3 STOCKIN-91725 52.84 − 7.1 
STOCKIN-86259 53.09 − 7.6 STOCKIN-88012 52.82 − 7.3 
STOCKIN-89039 53.08 − 8.2 STOCKIN-40363 52.77 − 6.2 
STOCKIN-58457 53.07 − 7.3 STOCKIN-16871 52.70 − 7.7 
STOCKIN-95708 53.05 − 7.6 STOCKIN-58337 52.59 − 7.5  
3.5. Protein-ligand interaction residues including Trp208, Ile228, Leu297, and Val330. Similarly, 
STOCKIN-89115 and STOCKIN-88290 which are structural derivatives 
The ligand-protein binding site interaction constitutes the basis of of 2-chromenone [46,70,71] formed hydrogen bonding interactions 
molecular recognition processes [50,68] and it underpins the discovery, with Arg295, Thr298, Tyr316 and Glu320. STOCKIN-88290 formed 
design and development of new therapeutic entities [69]. The docking hydrophobic contacts with Arg298, while STOCKIN-89115 formed with 
study enables the elucidation of small molecules with high specificity Ile296 and Leu297 (Table 3 and Fig. S1e). 
and affinity for the target. All the compounds were observed to dock at STOCKIN-76694 and STOCKIN-19207 formed hydrophobic contacts 
the S-adenosine-L-homocysteine (SAH) ligand binding cavity. The ex- with Met210, His226, Ile228, and Glu229. STOCKIN-76694 formed a 
amination of all the protein-ligand complexes (Fig. 6, S1a-i, and Table 3) hydrogen bond with Glu229 (Table 3 and Fig. S1f) while STOCKIN- 
revealed the residues Cys181, Trp208, His226, Ile228, Ile272, Tyr275, 19207 formed one with His226 (Table 3 and Fig. S1d). STOCKIN- 
and Val330 as critical for binding, similar to an earlier study [21]. The 95708 formed hydrogen bond interaction with Glu7, Arg227 and 
nature of interactions exhibited in all the complexes included pi-anion, Glu236 but STOCKIN-86259 did not form such interaction. However, 
pi-pi stacking, pi-alkyl, pi-sigma and hydrogen bonding. Out of the 10 hydrophobic interactions for STOCKIN-95708 and STOCKIN-86259, 
selected hits, STOCKIN-54848 formed four hydrogen bonds with active included residues Ala9, Leu13, Arg16, Asp218, Ala28, Tyr136, 
site residues (Fig. 6a) followed by STOCKIN-95708 (Fig. 6b) and Phe307,His226, Val308, Ala311, and Pro312 (Table 3 and Fig. S1g). 
STOCKIN-89115 (Fig. 6c), both forming three each with critical resi- STOCKIN-54848 had lower binding energy of − 10.1 kcal/mol compared 
dues. STOCKIN-68720, STOCKIN-44724, and STOCKIN-47277 are to 22,26-azasterol (7.6 kcal/mol). It formed four hydrogen bonds with 
structurally similar and analogues of chromenone [46]. While critical residues His226, Glu229, Tyr275 and Val331 and formed hy-
STOCKIN-44724 did not form hydrogen bonds with any residue in the drophobic contacts with Cys181, Lys225 and Ile228 (Table 3 and 
binding site of the target protein, compounds STOCKIN-68720 and Fig. 5a) while 22,26-azasterol had hydrophobic interactions with 
STOCKIN-47277 on the other hand formed one hydrogen bond each Phe100, Lys198 and Pro199 (Table 3 and Fig. S1i) [21]. One of the 
with Glu207 and Val331, respectively (Table 3 and Figs. S1a–c). Hy- known inhibitors of LdSMT, X2 with IC50 of 2.5 μM was found not to 
drophobic profiling revealed that STOCKIN-89115, STOCKIN-68720 form hydrogen bonds with any of the amino acids in the binding pocket 
and STOCKIN-44724 formed hydrophobic interactions with similar of the target. However, hydrophobic interactions were formed with 
Fig. 6. The 2D interactions of the LdSMT-hits complexes as visualized in Discovery Studio. (a) STOCKIN-54848, (b) STOCKIN-95708, (c) STOCKIN-89115, and (d) 
legend interactions. 
7
P.O. Sakyi et al.                                                                                                                                                                                   I n  f o  r m   a t i c  s  i n   M   e d  i c i n  e   U  n  l o c  k  e d 37 (2023) 101162
Table 3 
Binding energy and predicted binding site residues involved in the LdSMT-ligand complexes formed.  
Ligands Binding Energy/kcal/mol Binding residues 
Hydrogen bonds Hydrophobic bonds 
− 10.1 His226, Glu229, Tyr275, Val331 Cys181, Lys225, Ile228 
− 8.9 Arg295, Thr298, Tyr316 Ile296, Leu297 
− 8.7 Glu207 Trp208, Ile228, Val330 
− 8.4  Glu207, Trp208, Ile228, Glu229, Ile272, Val330 
− 8.4 Glu229 Tyr206, Trp208, Met210, Pro216, Ile228, Ser271, Ile272 
− 7.8 Thr298, Glu320 Arg289 
− 7.8 His226 Glu178, Cys181, Met210, Ile228, Glu229 
− 7.8 Val331 Trp208, Ile228, Tyr275 
− 7.6 Glu7, Arg227, Glu236 Ala9, Leu13, Arg16, Asp218, His226, Pro312 
− 7.6 Arg32, Asp35 Ala28, Tyr136, Phe307, Val308, Ala311 
(continued on next page) 
8
P.O. Sakyi et al.                                                                                                                                                                                   I n  f o  r m   a t i c  s  i n   M   e d  i c i n  e   U  n  l o c  k  e d 37 (2023) 101162
Table 3 (continued ) 
Ligands Binding Energy/kcal/mol Binding residues 
Hydrogen bonds Hydrophobic bonds 
− 7.0  Phe100, Gly200, Cys202, Tyr343, Ile344 
− 7.6 Glu102, Gly200 Phe100, Lys198, Pro199  
residues Phe100, Gly200, Cys202, Tyr343 and Ile344 (Table 3 and and it is evaluated to predict the bioavailability and biological activity of 
Fig. S1h). the hit compounds [78]. Recent in silico studies employed solubility 
predictions as criteria for the selection of hit compounds, as many drugs 
developed have failed clinical trials due to poor solubility [21,67,74]. 
3.6. Physicochemical properties of selected molecules The identified potential hit compounds (Table 4) were predicted to be 
soluble suggesting their bioavailability potential. In addition, bioavail-
The Lipinski’s rule of five (Ro5) was used to determine the physi- ability score (BS) which is the fraction of an administered drug that 
cochemical profiling of the compounds [72,73]. The Ro5 stipulates that reaches systemic circulation was assessed to differentiate well absorbed 
for a compound to be orally active then certain parameters must be met compounds from the poor. The selected hit compounds as well as the 
which include molecular weight (Mw) ≤ 500 Da, LogP ≤5, hydrogen known inhibitors were predicted to have a bioavailability score of 0.55. 
bond donors ≤5, hydrogen bond acceptors ≤10 and 40 ≤ molar A de novo design of potential inhibitors of LdSMT identified the novel 
refractivity ≤140 [73]. The Ro5 was used as a criterion for selecting ligands to possess a bioavailability score of 0.55 [21]. A similar in silico 
orally active compounds [65]. Similarly, other studies have also used study found natural products compounds to have a bioavailability score 
Ro5 in selecting compounds from natural products sources with the of 0.55 [74]. 
potential to be orally active [74–76]. All the 10 selected hit compounds Molar refractivity (MR) is the measure of the total polarity of a drug 
(Table 4) complied with Ro5 and hence predicted to possess good and it is estimated to provide useful information concerning the phar-
physicochemical properties. The known inhibitors, azetimibe, imipra- macokinetics and pharmacodynamics of the molecule [79,80]. The 
mine, X2 and 22,26-azasterol were also predicted to be orally active acceptable MR range for drug-like molecules is 40–130 [21]. The pre-
since they did not violate any of the Ro5. The Veber’s rule suggests dicted hit compounds (Table 4) were all found to have overall polarity 
compounds can be orally active if the number of rotatable bonds ≤10 within the acceptable range. Similar predictions were observed for 22, 
and polar surface area ≤140 Å2 [77]. Both the selected hit compounds 26-azasterol, X2, imipraminn and azetimibe. Synthetic accessibility 
and known inhibitors were predicted to possess less than 10 rotatable (SA) is the measure of the overall synthetic feasibility of a molecule and 
bonds and polar surface areas less than 140 Å2, suggesting their po- drug-like molecules are predicted to have a synthetic accessibility score 
tential as orally active compounds. lower than 6 [81]. Apart from X2 which showed SA score close to 6 (that 
Drug solubility is the maximum concentration of the drug that can is 5.65), the selected inhibitors, 22,26-azasterol, imipramine and 
completely dissolve in a solvent at a constant temperature and pressure 
Table 4 
Predicted physicochemical and pharmacological profiles of selected compounds and known inhibitors of LdSMT as well as some known drugs for leishmaniasis 
treatment.  
Compounds Mw (g/mol) NRB MR TPSA (Å2) LogS GI vRo5 BS SA BBB P-gp 
STOCKIN-54848 447.44 5 126.78 132.88 − 1.26 High 0 0.55 4.51 No No 
STOCKIN-89115 358.43 5 104.73 80.57 − 3.56 High 0 0.55 3.29 No Yes 
STOCKIN-68720 339.39 4 98..88 62.91 − 4.04 High 0 0.55 3.32 Yes No 
STOCKIN-44724 325.36 4 93.91 62.91 − 3.74 High 0 0.55 3.15 Yes No 
STOCKIN-76694 325.36 6 89.69 83.40 − 3.64 High 0 0.55 3.69 No Yes 
STOCKIN-88290 386.48 5 98.88 62.91 − 4.27 High 0 0.55 3.95 No Yes 
STOCKIN-19207 467.47 5 127.27 115.25 − 4.26 High 0 0.55 4.26 No Yes 
STOCKIN-47277 383.44 7 110.02 72.14 − 4.28 High 0 0.55 3.62 Yes No 
STOCKIN-95708 471.89 5 121.61 132.47 − 5.01 High 0 0.55 4.93 No Yes 
STOCKIN-86259 344.58 7 104.58 91.73 − 3.27 High 0 0.55 3.54 No Yes 
Imipramine 280.41 4 93.76 6.48 − 4.76 High 0 0.55 2.99 Yes No 
Azetimibe 409.43 6 112.97 60.77 − 4.92 High 1 0.55 3.37 Yes Yes 
X2 441.69 5 138.05 29.54 − 5.47 High 1 0.55 5.65 No No 
22,26-Azasterol 403..64 2 125.08 52.49 − 5.66 High 1 0.55 4.84 Yes Yes  
9
P.O. Sakyi et al.                                                                                                                                                                                   I n  f o  r m   a t i c  s  i n   M   e d  i c i n  e   U  n  l o c  k  e d 37 (2023) 101162
azetimibe were predicted to possess SA score less than 5 (Table 4). 3.8. Biological activity of selected hits 
Overall, the selected hit compounds were predicted to be orally active 
since they conform to the Ro5. The Open Bayesian machine learning algorithm, Prediction of Ac-
tivity Spectra for Substances (PASS) [44] was used to predict the bio-
3.7. Pharmacological and toxicity profiling of selected molecules logical activity of the selected molecules (Table 5). PASS queries are 
based on the structure activity relationship (SAR) analysis of the 
Pharmacokinetics deals with the movement of a drug in the body chemical entities present in the database [89]. 
[82,83] and its prediction is essential as drug candidates must reach A total of 7 compounds comprising STOCKIN-54848, STOCKIN- 
their targets to elicit biological responses [83]. Absorption, distribution, 89115, STOCKIN-68720, STOCKIN-44724, STOCKIN-76694, STOCKIN- 
metabolism and excretion (ADME) predictions are employed as criteria 47277, and STOCKIN-95708 were predicted to possess antileishmanial 
for filtering chemical libraries in the identification of hits [46,84]. Some properties with predicted probability of activity (Pa) values of 0.256, 
of the key parameters computed herein include gastrointestinal ab- 0.226, 0.391, 0.421, 0.372, 0.479 and 0.380; and probability of inac-
sorption (GI), blood-brain barrier (BBB) and permeability glycoprotein tivity (Pi) values of 0.096, 0.166, 0.052, 0.041, 0.061, 0.027 and 0.057, 
(P-gp). A high GI means the administered drugs can easily be absorbed respectively. The higher Pa values compared to the Pi, suggest the 
in the intestines [85]. SwissADME [42] predicted all 10 hit compounds likelihood of these compounds to inhibit the LdSMT and hence worthy of 
to possess high GI meaning they can easily be absorbed in the intestines experimental evaluation [90]. In addition, all the compounds were 
and transported to their target receptors. The known inhibitors, X2, predicted to have antineoplastic effects with Pa greater than Pi. Previous 
imipramine, azetimibe and 22,26-azasterol were also predicted to studies have identified a number of anticancer compounds (sunitinib, 
possess a high GI. Interestingly, an earlier study predicted the anti- sorafenib and lapatanib) which have been evaluated against Leishmania 
leishmanial drugs amphotericin b, paromomycin and miltefosine to parasites with some level of successes of which some including bexar-
show low GI which suggest their low absorption into the blood stream otene, imatinib and tamoxifen are already in clinical trials for leish-
[66]. A drug can cross the brain-barrier to the brain and attach itself to maniasis treatment [66,91–94]. 
specific receptors to elicit activation of certain signal pathways [86]. STOCKIN-89115 and STOCKIN-88290 were predicted to possess 
Seven of the selected molecules comprising STOCKIN-54848, dermatological properties with Pa values of 0.196 and 0.244 and Pi 
STOCKIN-89115, STOCKIN-76694, STOCKIN-88290, STOCKIN-19207, values of 0.172 and 0.123, respectively (Table 5) suggesting they can be 
STOCKIN-95708 and STOCKIN-86259 (Table 4) were predicted not to explored for post kala-azar leishmaniasis [66,95]. STOCKIN-54848, 
cross the brain-barrier. However, STOCKIN-68720, STOCKIN-44714 STOCKIN-68720, STOCKIN-44724, STOCKIN-47277 and 
and STOCKIN-47277 were predicted to have the ability of crossing the STOCKIN-86259 were also predicted as 3′-dimethylstaurosporine O- 
BBB. Compared to the known inhibitors, apart from X2 which was methyltransferase inhibitors with Pa of 0.260, 0.312, 0.28, 0.274 and 
predicted not to cross the BBB, the rest comprising imipramine, azeti- 0.208; and Pi of 0.081, 0.031, 0.038, 0.04 and 0.104, respectively. In 
mibe and 22,26 show BBB permeation. A recent study to identify po- addition, STOCKIN-47277 was predicted as an inhibitor of phenol 
tential leishmanicides targeting cell division cycle (cdc) 2-related kinase O-methyltransferase with Pa of 0.23 and Pi of 0.16. Both 3′-dimethyl-
12 (CRK12) shows that five of the selected hit compounds possessed BBB staurosporine O- methyltransferase and phenol O-methyltransferase 
permeation [87]. A drug which is an inhibitor of P-gp will possibly have belong to the family of transfer of one-carbon group methyltransferases 
an increased bioavailability at the site of activity [88]. While [96–98] of which the target protein, LdSMT is also a member. Therefore, 
STOCKIN-89115, STOCKIN-76694, STOCKIN-88290, STOCKIN-19207, the predicted inhibition of 3′-dimethylstaurosporine O- methyl-
STOCKIN-95708 and STOCKIN-86259 were predicted to be substrates of transferase and phenol O-methyltransferase by the selected hits suggest 
P-gp, STOCKIN-54848, STOCKIN-68720, STOCKIN-44724 and the compounds to possess the potential of suppressing ergosterol 
STOCKIN-47277 were otherwise. On the other hand, four known in- biosynthesis by inhibiting LdSMT. Similarly, these predictions are 
hibitors were predicted as P-gp substrates except X2 and imipramine. corroborated by an earlier study which identified potential LdSMT in-
The toxicity profiles of the 10 compounds and the known drugs were hibitors via de novo design. The identified hits were predicted to suppress 
predicted using OSIRIS Data Warrior 5.0.0 [43]. STOCKIN-54848 and 3′-dimethylstaurosporine O- methyltransferase and phenol O- methyl-
STOCKIN-19207 were predicted to be mutagenic while STOCKIN-54848 transferase [21]. Furthermore, STOCKIN-89115, STOCKIN-68720, 
and STOCKIN-95708 were predicted as tumorigenic. STOCKIN-88290 STOCKIN-44724 and STOCKIN-47277 (Table 5) were predicted to be 
and STOCKIN-47277 were predicted to possess reproductive and irri- membrane permeability inhibitors with Pa values of 0.394, 0.359, 0.550 
tant effects, respectively. This notwithstanding, optimization of the and 0.541; and Pi values of 0.206, 0.222, 0.117 and 0.222, respectively. 
potential hit compounds to generate analogues with improved toxicities Suramin and other antibiotics (such as fosfomycin, cycloserine, vanco-
can be generated for their effective use in finding a lasting solution for mycin and bacitracin) used in treating African trypanosomiasis, cancer 
leishmaniasis treatment. and bacterial infections exhibit their mode of action by suppressing 
membrane permeability [99–101]. This suggests the compounds can be 
investigated for their ability to inhibit membrane channels and protect 
Table 5 
Biological activity prediction of selected hits obtained from pharmacophore-based drug design.  
Selected Hits Antileishmanial activities Antineoplastic activities Dermatologic Membrane Permeability Inhibitor 
Pa Pi Pa Pi Pa Pi Pa Pi 
STOCKIN-54848 0.256 0.096 0.214 0.104 – – – – 
STOCKIN-89115 0.226 0.166 0.334 0.133 0.196 0.172 0.394 0.206 
STOCKIN-68720 0.391 0.052 0.391 0.108 – – 0.359 0.222 
STOCKIN-44724 0.421 0.041 0.373 0.115 – – 0.55 0.117 
STOCKIN-76694 0.372 0.061 0.286 0.102 – – – – 
STOCKIN-88290 – – 0.349 0.126 0.244 0.123 – – 
STOCKIN-19207 – – 0.264 0.123 – – – – 
STOCKIN-47277 0.479 0.027 0.391 0.108 – – 0.541 0.124 
STOCKIN-95708 0.380 0.057 0.252 0.184 – – – – 
STOCKIN-86259 – – 0.358 0.008 – – – –  
10
P.O. Sakyi et al.                                                                                                                                                                                   I n  f o  r m   a t i c  s  i n   M   e d  i c i n  e   U  n  l o c  k  e d 37 (2023) 101162
cells against toxic and hypotonic stress that will lead to disease exac- therefore used potential novel scaffolds (STOCKIN-54848, 
erbation. Altogether, the selected hits possess novel scaffolds with the STOCKIN-76694, STOCKIN-95708) without chromenone backbone 
potential of inhibiting LdSMT as therapeutic moieties. together with the known drug (22,26-azasterol) and inhibitor (X2) for the 
MD simulations. Compound STOCKIN-54848 had four hydrogen bonds 
coupled with its low binding energy (− 10.1 kcal/mol) making it an 
3.9. Selection of hits for molecular dynamics simulation interesting candidate to investigate further using MD simulation and 
MM-PBSA calculations. 
Hit compounds with predicted high Pa pertaining to relevant biological 
activity require experimental validation [102–105]. STOCKIN-54848, 
STOCKIN-89115, STOCKIN-68720, STOCKIN-44724, STOCKIN-76694, 3.10. Molecular dynamics simulation 
STOCKIN-47277, and STOCKIN-95708 were predicted to possess anti-
leishmanial properties with Pa > Pi. The medicinal properties for chro- A 100 ns MD simulation was performed to check the relative stability 
menone and its derivatives for treatments of various human ailments have of the LdSMT-ligand complexes. The output files were used to generate 
been explored [106]. Chromenone scaffolds with promising anti- root mean square deviation (RMSD), root mean square fluctuation 
leishmanial potentials include encecalol angelate which inhibited axenic (RMSF), and radius of gyration (Rg) plots of the complexes for analysis. 
amastigotes of L. donovani at an IC50 value of 14.6 μg/mL [70]. The com- The RMSD of the protein backbone of each complex were calculated and 
pound 2-(6,8 -Dichloro-2-methyl-4H chromene-4-ylidene) malono plotted against time (Fig. 7a). The RMSD plot explains the fluctuations of 
nitrile and its analogue, 2-Amino-3-(6,8-dichloro-2-methyl-4H the LdSMT protein backbone and evaluates the stability of the protein- 
chromen-4-ylidene)prop-1-ene-1,1,3-tricarbonitrile also stopped the ligand complexes [84]. From the RMSD trajectories, it was observed 
growth of intracellular forms at 0.58 and 0.59 μg/mL, respectively [107]. A that both the unbound protein as well as complexes of STOCKIN-54848, 
recent in silico studies to find putative hits against L. major identified STOCKIN-76694, STOCKIN-95708, X2 and 22,26-azasterol were equil-
chromen-2-one derivatives as potential inhibitors of trypanothione reductase ibrated with RMSD within the range of 0 and 0.6 nm (Fig. 7a), consistent 
corroborating the earlier in vitro and in vivo studies [46]. Based on these with the de novo design of inhibitors against the target [21]. The RMSD 
findings, the selected hits with chromenone moieties including of the unbound protein rose from 0 nm to 0.46 nm within a period of 20 
STOCKIN-89115, STOCKIN-68720, STOCKIN-44724, STOCKIN-88290 and ns and then plateaued to the end of the 100 ns simulation. Similar trend 
STOCKIN-47277 were not considered for MD simulation. The study was observed for all the other complexes. On comparing the trajectories, 
Fig. 7. RMSD, RMSF, Rg and SASA trajectories obtained from the molecular dynamic simulations. (a) RMSD versus time, (b) RMSF versus residue number, and (c) Rg 
versus time, and (d) SASA versus time for protein-ligand complexes including X2 and 22,26-Azasterol. 
11
P.O. Sakyi et al.                                                                                                                                                                                   I n  f o  r m   a t i c  s  i n   M   e d  i c i n  e   U  n  l o c  k  e d 37 (2023) 101162
the unbound protein trajectory achieved stability with the highest RMSD thereafter. Similar trajectory was observed for X2 and 22,26-azasterol 
implying the complexes achieved stability upon equilibration. Among with stabilization of the trajectory occurring after 30 and 60 ns, 
the trajectories of the complexes, both STOCKIN-54848 and respectively during the 100 ns simulation period. However, the com-
STOCKIN-95708 recorded the highest RMSD averaging about 0.2948 plexes of STOCKIN-54848 and STOCKIN-76694 showed an initial 
and 0.2894 nm, respectively (Table 6) compared to the X2 (0.2515) and decline of SASA in the first 20 ns from 183 to 187 nm2, respectively to 
22,26-azasterol (0.2617). Furthermore, STOCKIN-76694 had an average ~160 nm2 before rising for the next 10 ns. While STOCKIN-76694 sta-
RMSD of 0.2529, lower than 22,26-azasterol but comparable to X2 in bilized during the remaining 70 ns, the SASA for STOCKIN-54848 
terms of stability among the compounds [108]. decreased to 155 nm2 and then remained stable for the remaining 
The amino acid residues in the active site are critical in the stability period of the simulation. On the other hand, the SASA of the STOCKIN- 
of protein-ligand complexes [109]. RMSF was used to measure the 95708 decreased from 180 to 165 nm2 in the initial 40 ns of the simu-
fluctuations of the amino acid residues to the stability of the complex lation and then rose to 175 nm2 in the next 20 ns. It decreased again to 
(Fig. 7b). Amino acid residues between regions 15–100, 200–250, and 155 nm2 at the 90 ns mark before becoming stable. Comparatively, the 
280–320 were identified to be critical for the protein-ligand complexes average SASA recorded for STOCKIN-54848, STOCKIN-76694, 
[21]. In this study, amino acid residues within these same regions were STOCKIN-95708, X2 and 22,26-azasterol are 166.152, 172.178, 
found to fluctuate for all the compounds implying that they are critical 171.837, 163.254 and 167.139 nm2, respectively (Table 6). As a rule of 
for binding. This notwithstanding, the highest fluctuation was observed thumb, low SASA values imply limited solvent surface charges resulting 
around the region of 280–310 consistent with other studies [21,58,110]. in structural restrictions of protein-ligand complexes leading to a more 
The intermolecular forces of attraction between ligands and proteins compact and stable configuration [111]. Consequently, considering the 
determine the rigidness of the complex with a strong force of interaction average of the computed SASA values for the various complexes, the 
depicting very compact complex and vice versa [50]. The compactness order of stability is accordingly X2 > STOCKIN-54848 > 22,26-azas-
of the protein-ligand complexes were analyzed by measuring the Rg [66, terol > STOCKIN-95708 > STOCKIN-76694. The low SASA values 
111]. The Rg trajectory (Fig. 7c) ranging between 1.9 and 2.075 nm recorded for all the complexes signify their stability throughout the 100 
suggested the rigidness of the complexes during the entire simulation ns simulation period. 
period. The Rg of the unbound LdSMT declined from 2.05 to about 1.95 
nm over a period of 40 ns, then became steady with an average Rg of 3.11. MM-PBSA calculations 
1.955 nm until about 80 ns, and then rose steadily to 1.975 nm. The Rg 
of LdSMT-STOCKIN-54848 complex decreased steadily from 2.04 to The MM-PBSA method was used in calculating the free energy of 
about 1.96 nm for 20 ns and then increased slightly to 2.10 nm for 40 ns binding (ΔGbind), van der Waals interactions energy (ΔGvdW), electro-
and then declined steadily thereafter for the next 40 ns. The Rg of static energy (ΔGele), polar solvation energy (ΔGpol, sol) and non-polar 
LdSMT-STOCKIN-76694 and LdSMT-STOCKIN-95708 complexes expe- solvation energy (ΔGSASA) of the selected protein-ligand complexes 
rienced similar trends with gradual decline from 2.05 to 1.975 nm over a [112,113]. A high negative ΔGbind value for each complex except for 
period of about 50 ns. The Rg values then rose to about 2.025 nm for the LdSMT-STOCKIN-95708 complex signifies a strong affinity of the small 
next 20 ns and then decreased steadily to an average of 2.0 nm for the molecules towards the LdSMT corroborating results from the docking 
last 30 ns. Similar fluctuation trends were observed for LdSMT-X2 and studies [50,114]. The complex of STOCKIN-54848 which was predicted 
LdSMT-22,26-azasterol complexes with LdSMT-X2 showing the lowest to have the least binding energy (− 10.1 kcal/mol) from AutoDock Vina 
Rg among all the complexes. The average Rg for STOCKIN-54848, was shown to have free binding energy (− 127.838 kJ/mol) lower and 
STOCKIN-76694, STOCKIN-95708, X2 and 22,26-azasterol complexes comparable to the known inhibitors, 22,26-azasterol (− 72.305 kJ/mol) 
were found to be 1.9900, 2.0091, 2.0075, 1.9669 and 1.9992 nm, and X2 (− 129.725 kJ/mol), respectively. Similarly, STOCKIN-76694 
respectively, consistent with a previous study [21]. This further suggests exhibited the lowest free binding energy (− 149.899 kJ/mol) among 
that X2 formed a stable and rigid complex with LdSMT followed by the selected hits as well as X2 and 22,26-azasterol. STOCKIN-95708 
STOCKIN-54848, 22,26-azasterol, STOCKIN-95708 and showed a high free binding energy of +4190.558 kJ/mol when it had 
STOCKIN-76694. The average Rg value of the five systems was around − 7.6 kcal/mol obtained from Autodock Vina rendering STOCKIN-95708 
1.9945 nm suggesting that all the complexes retained their rigidity not fit to be a hit. This notwithstanding, the main contributor to the free 
throughout the simulations and formed stable complexes with LdSMT. energy of binding was found to be van der Waal (ΔGvdW) and is 
Moreover, to observe the effects of solvent accessibility occasioned responsible for the embedded hydrophobic binding site of the target 
by loose packing on the stability of the protein-ligand complexes, SASA protein. The next dominating energy contributor is ΔGele with values 
was computed and plotted. The protein-ligand complexes and the un- ranging from − 94.983 to − 3.390 kJ/mol followed by polar solvation 
bound protein displayed similar SASA trajectory within the range of energies with − 15.475 kJ/mol being the least energy contributed for the 
140–190 nm2 (Fig. 7d) The SASA for the unbound LdSMT decreased complexes (Table 7). The energy contributions suggest that the selected 
from 180 nm2 to 150 nm2 within the first 30 ns after which it rose hits except STOCKIN-95708 have more propensity to be potential in-
gradually to 160 nm2 for the next 10 ns and then remained stable hibitors of LdSMT. 
Table 6 
The maximum, minimum and average RMSD, RMSF and Rg values obtained from MD simulations.  
Complex STOCKIN-54848 STOCKIN-76694 STOCKIN-95708 X2 22,26-Azasterol 
RMSD (nm) Maximum 0.5890 0.5052 0.5783 0.5024 0.5229 
Minimum 0.0005 0.0005 0.0005 0.0005 0.0005 
Average 0.2948 0.2529 0.2894 0.2515 0.2617 
RMSF (nm) Maximum 0.8140 0.7223 0.8001 0.6327 0.5129 
Minimum 0.0818 0.0867 0.0710 0.0685 0.0719 
Average 0.4479 0.4045 0.4355 0.3506 0.2920 
Rg (nm) Maximum 2.0374 2.0508 2.0526 2.0131 2.0662 
Minimum 1.9426 1.9674 1.9623 1.9206 1.9323 
Average 1.9900 2.0091 2.0075 1.9669 1.9992 
SASA (nm2) Maximum 183.370 187.281 189.524 180.466 187.260 
Minimum 148.934 157.075 154.149 146.042 147.017 
Average 166.152 172.178 171.837 163.254 167.139  
12
P.O. Sakyi et al.                                                                                                                                                                                   I n  f o  r m   a t i c  s  i n   M   e d  i c i n  e   U  n  l o c  k  e d 37 (2023) 101162
The energy contributions of critical amino acid residues are essential 
for the stability of protein-ligand complexes [38,115]. The per-residue 
decomposition was employed to determine the energy input of each 
residue and as a rule of thumb, residues with energies ≥+5 kJ/mol or ≤
− 5 kJ/mol are regarded as critical for ligand binding [116]. Among the 
three selected hits, STOCKIN-76694 had residues Trp208, Glu229 and 
Leu230 (Fig. 8) contributing energies above or below +5 or − 5 kJ/mol. 
The residues Thr8, Pro10, Asn18 and Lys19 (Fig. S2a) were found to 
provide energies lower than − 5 kJ/mol for the stability of the 
STOCKIN-54848 complex. Compared to the known inhibitors, while 
none of the residues (Fig. S2b) contributed energies ≥ +5 kJ/mol or ≤
− 5 kJ/mol for the stability of X2, that of Glu102 and Lys198 (Fig. S2c) 
were critical for the stability of the 22,26-azasterol complex. While the 
earlier study identified Tyr92 and Ala96 to be critical residues for ligand 
binding to LdSMT target [21], the current work shows that Thr8, Pro10, 
Asn18, Lys19, Glu102, Lys198, Trp208, Glu229 and Leu230 amino acids 
within the binding site also contribute plausible energies required for 
the stability of the protein-ligand complexes. 
4. Future outlook 
Fig. 8. Per-residue energy decomposition trajectory for the LdSMT-STOCKIN- 
Pharmacophore-based drug design with its added advantage of 76694 complex. 
incorporating diversity of features and promoting specificity plays 
pivotal role in the identification of novel chemotypes for different pro- antileishmanial potentials. 
tein targets [67,117–119]. Integration of the predictions with experi- Furthermore, the coordination of the predicted compounds with 
mental validation will ensure a better means of evaluating the structure, transition metals targeting sterol methyltransferase could provide a 
function and modulation of essential therapeutic targets needed to plausible means of combating leishmaniasis. 
accelerate the drug design pipeline [119–121]. Therefore, the predicted 
hits in this study could be validated by the efficient and reliable 5. Conclusion 
experimented assay to confirm their potential as lead molecules. 
Mechanisms of binding of the identified hits could be elucidated via The present work employed pharmacophore-based design, molecu-
thermal melt assays [122] followed by evaluation of the physiological lar docking and dynamics simulations studies to predict hits with unique 
activities [123] against the Leishmania parasite. Furthermore, a good scaffolds as potential inhibitors of LdSMT for leishmaniasis treatment. In 
understanding of the inhibitory effect of the hit molecules would be all, 10 compounds were predicted as potential inhibitors of Leishmania 
achieved by performing experimental studies comprising the kinetic and parasites attenuating LdSMT with binding energies between − 7.6 and 
thermodynamic studies [124] of the protein-hit residence time. − 10.1 kcal/mol, which were lower or comparable to the known inhib-
The recent upsurge in the interest of the concept of computer-aided itor, 22,26-azasterol. MD simulations and MM-PBSA corroborated the 
drug design (CADD) is attributed to the shortened duration, higher docking results with Trp208 and Val330 identified as novel residues 
selectivity and reduction in the cost of drug design compared to the critical for ligand binding. While all the compounds were predicted to 
traditional means of drug design and development [125]. Some drugs possess antineoplastic properties, seven of the hits (STOCKIN-54848, 
currently in use such as saquinavir, a drug for the treatment of AIDS STOCKIN-89115, STOCKIN-68720, STOCKIN-44724, STOCKIN-76694, 
targeting proteases of HIV 1 and HIV 2, captopril for treating high blood STOCKIN-47277 and STOCKIN-95708) were also predicted to exhibit 
pressure by inhibiting the conversion of angiotensin-converting enzyme antileishmanial effects. The bioactivity predictions, pharmacological 
and zanamivir which target neuraminidase for influenza A and B profiles, binding energies, and binding mechanisms suggest these hits as 
treatments were designed by CADD [126]. An in silico study identified potential inhibitors of the target protein which need experimental 
rutin, a flavonol known for its antioxidant, anticarcinogenic and car- validation to confirm their antileishmanial activity. 
dioprotective activities to possess antileishmanial properties targeting 
Leishmania donovani 3-mercaptopyruvate sulfurtransferase [127,128]. Authors’ contributions: conceptualization of the project 
In vitro studies carried out shows that rutin inhibits Leishmania donovani 
promastigotes and amastigotes with an IC of 40.95 and 90.09 μM, The project idea was born by P.O.S., S.K.K. and R.K.A. P.O.S. 50 
respectively [128]. Similarly, an in silico study identified bisindolylma- designed the project and executed all the computational analysis with 
leimide IX (BIM IX) to possess multitargeting potential for the treatment insightful contributions from S.K.K, R.K.A., E.B., W.A.M., and M.D.W. 
of SARS-CoV-2 [63]. In vitro studies validated the docking results with All the authors made their inputs and was accepted by all after P.O.S 
IC50 of 28.18 μM targeting 3CLpro [63]. Based on the aforementioned, wrote the first draft before manuscript submission. 
biological activity and cytotoxicity testing via in vitro and in vivo analysis 
can be carried out on the identified potential hits to assess their 
Table 7 
MM-PBSA free binding energy assessment scores of selected molecules.  
Complex ΔGvdW (kJ/mol) ΔGele (kJ/mol) ΔGpol,sol (kJ/mol) ΔGSASA (kJ/mol) ΔGbind (kJ/mol) 
STOCKIN-54848 − 181.010 ± 17.506 − 94.983 ± 26.865 162.483 ± 37.132 − 14.329 ± 2.329 − 127.838 ± 25.330 
STOCKIN-76694 − 217.243 ± 85.589 − 6.439 ± 10.576 90.570 ± 47.465 − 16.786 ± 6.646 − 149.899 ± 58.600 
STOCKIN-95708 4135.241 ± 1670.883 − 43.493 ± 19.307 118.716 ± 47.409 − 19.905 ± 8.383 4190.558 ± 1688.261 
X2 − 142.275 ± 50.679 − 3.390 ± 6.417 29.163 ± 21.101 − 13.223 ± 4.385 − 129.725 ± 51.799 
22,26-Azasterol − 0.047 ± 0.042 − 56.829 ± 37.192 − 15.475 ± 37.519 0.045 ± 2.716 − 72.305 ± 59.057  
13
P.O. Sakyi et al.                                                                                                                                                                                   I n  f o  r m   a t i c  s  i n   M   e d  i c i n  e   U  n  l o c  k  e d 37 (2023) 101162
Funding [14] Lorente SO, Rodrigues JCF, Jiménez CJ, Joyce-Menekse M, Rodrigues C, Croft SL, 
Yardley V, De Luca-Fradley K, Ruiz-Pérez LM, Urbina J, De Souza W, González 
Pacanowska D, Gilbert IH. Novel azasterols as potential agents for treatment of 
This research received no external funding leishmaniasis and trypanosomiasis. Antimicrob Agents Chemother 2004;48: 
2937–50. https://doi.org/10.1128/AAC.48.8.2937-2950.2004. 
Institutional review board statement [15] Gros L, Castillo-Acosta VM, Jiménez CJ, Sealey-Cardona M, Vargas S, Estévez AM, Yardley V, Rattray L, Croft SL, Ruiz-Perez LM, Urbina JA, Gilbert IH, González- 
Pacanowska D. New azasterols against Trypanosoma brucei: role of 24-sterol 
Not applicable. methyltransferase in inhibitor action. Antimicrob Agents Chemother 2006;50: 
2595–601. https://doi.org/10.1128/AAC.01508-05. 
[16] Rodrigues JCF, Attias M, Rodriguez C, Urbina JA, De Souza W. Ultrastructural 
Informed consent statement and biochemical alterations induced by 22,26-azasterol, a Δ24(25)-sterol 
methyltransferase inhibitor, on promastigote and amastigote forms of Leishmania 
Not applicable. amazonensis. Antimicrob Agents Chemother 2002;46:487–99. https://doi.org/ 10.1128/AAC.46.2.487-499.2002. 
[17] Magaraci F, Jimenez Jimenez C, Rodrigues C, Rodrigues JCF, Vianna Braga M, 
Declaration of competing interest Yardley V, De Luca-Fradley K, Croft SL, De Souza W, Ruiz-Perez LM, Urbina J, Gonzalez Pacanowska D, Gilbert IH. Azasterols as inhibitors of sterol 24-meth-
yltransferase in leishmania species and trypanosoma cruzi. J Med Chem 2003;46: 
The authors declare that they have no known competing financial 4714–27. https://doi.org/10.1021/jm021114j. 
interests or personal relationships that could have appeared to influence [18] Rodrigues JCF, Bernardes CF, Visbal G, Urbina JA, Vercesi AE, de Souza W. Sterol methenyl transferase inhibitors alter the ultrastructure and function of the 
the work reported in this paper. leishmania amazonensis mitochondrion leading to potent growth inhibition. 
Protist 2007;158:447–56. https://doi.org/10.1016/j.protis.2007.05.004. 
Acknowledgments [19] Sakyi PO, Amewu RK, Devine RNOA, Ismaila E, Miller WA, Kwofie SK. The search for putative hits in combating leishmaniasis: the contributions of natural products 
over the last decade. Nat Products Bioprospect 2021;11:489–544. https://doi. 
The authors are grateful to Ghana National Petroleum Corporation org/10.1007/S13659-021-00311-2. 
(GNPC) for the scholarship awarded Patrick Opare Sakyi for his PhD [20] Ghorbani M, Farhoudi R. Leishmaniasis in humans: drug or vaccine therapy? Drug Des Dev Ther 2018;12:25. https://doi.org/10.2147/DDDT.S146521. 
studies. We also acknowledge West African Centre for Cell Biology of [21] Sakyi PO, Broni E, Amewu RK, Miller WAI, Wilson MD, Kwofie SK. Homology 
Infectious Pathogens (WACCBIP), University of Ghana for granting free modeling, de Novo design of ligands, and molecular docking identify potential 
access to Zuputo, a DELL high performance computing system which inhibitors of leishmania donovani 24-sterol methyltransferase. Front Cell Infect Microbiol 2022:657. https://doi.org/10.3389/FCIMB.2022.859981. 0. 
helped immensely to the successful completion of this work. [22] Qing X, Lee XY, De Raeymaeker J, Tame JR, Zhang KY, De Maeyer M, Voet AR. 
Pharmacophore modeling: advances, limitations, and current utility in drug 
Appendix A. Supplementary data discovery. J Recept Ligand Channel Res 2014;7:81–92. https://doi.org/10.2147/ JRLCR.S46843. 
[23] Kaserer T, Beck KR, Akram M, Odermatt A, Schuster D, Willett P. Pharmacophore 
Supplementary data to this article can be found online at https://doi. models and pharmacophore-based virtual screening: concepts and applications 
org/10.1016/j.imu.2023.101162. exemplified on hydroxysteroid dehydrogenases. Molecules 2015;20:22799–832. https://doi.org/10.3390/molecules201219880. 
[24] Pandit D, So SS, Sun H. Enhancing specificity and sensitivity of pharmacophore- 
References based virtual screening by incorporating chemical and shape features - a case 
study of HIV protease inhibitors. J Chem Inf Model 2006;46:1236–44. https:// 
[1] Jones CM, Welburn SC. Leishmaniasis beyond East Africa. Front Vet Sci 2021;8: doi.org/10.1021/CI050511A. 
618766/1-618766/11. https://doi.org/10.3389/FVETS.2021.618766/BIBTEX. [25] Kumari A, Singh RK. Morpholine as ubiquitous pharmacophore in medicinal 
[2] Ready PD. Epidemiology of visceral leishmaniasis. Clin Epidemiol 2014;6: chemistry: deep insight into the structure-activity relationship (SAR). Bioorg 
147–54. https://doi.org/10.2147/CLEP.S44267. Chem 2020;96:103578/1-103578/78. https://doi.org/10.1016/J. 
[3] Celesia BM, Cacopardo B, Massimino D, Gussio M, Tosto S, Nunnari G, BIOORG.2020.103578. 
Pinzone MR. Atypical presentation of PKDL due to leishmania infantum in an [26] Thomas MP, Potter BVL. Discovery and development of the aryl O-sulfamate 
HIV-infected patient with relapsing visceral leishmaniasis. Case Rep. Infect. Dis. pharmacophore for oncology and womens health. J Med Chem 2015;58:7634–58. 
2014;2014:1 3. https://doi.org/10.1155/2014/370286. https://doi.org/10.1021/ACS.JMEDCHEM.5B00386. –
[4] Bilgic-Temel A, Murrell DF, Uzun S. Cutaneous leishmaniasis: a neglected [27] Amewu RK, Sakyi PO, Osei-safo D. Synthetic and naturally occurring heterocyclic 
disfiguring disease for women. Int. J. Women’s Dermatology. 2019;5:158–65. anticancer compounds with multiple biological targets. Molecules 2021;26:7134/ 
https://doi.org/10.1016/J.IJWD.2019.01.002. 1-7134/48. 
[5] Okwor I, Uzonna J. Social and economic burden of human leishmaniasis. Am J [28] Newman DJ, Cragg GM. Natural products as sources of new drugs from 1981 to 
Trop Med Hyg 2016;94:489–93. https://doi.org/10.4269/ajtmh.15-0408. 2014. J Nat Prod 2016;79:629–61. https://doi.org/10.1021/ACS. 
[6] Leta S, Dao THT, Mesele F, Alemayehu G. Visceral leishmaniasis in Ethiopia: an JNATPROD.5B01055. 
evolving disease. PLoS Neglected Trop Dis 2014;8:e3131/1-e3131/7. https://doi. [29] Wolber G, Langer T. LigandScout: 3-D pharmacophores derived from protein- 
org/10.1371/journal.pntd.0003131. bound ligands and their use as virtual screening filters. J Chem Inf Model 2005; 
[7] Sakyi PO, Amewu RK, Devine RNOA, Bienibuor AK, Miller WA, Kwofie SK. 45:160–9. https://doi.org/10.1021/CI049885E. 
Unravelling the myth surrounding sterol biosynthesis as plausible target for drug [30] Eswar N, Eramian D, Webb B, Shen MY, Sali A. Protein structure modeling with 
design against leishmaniasis. J Parasit Dis 2021;45:1152–71. https://doi.org/ MODELLER. Methods Mol Biol 2008;426:145–59. https://doi.org/10.1007/978- 
10.1007/S12639-021-01390-1. 1-60327-058-8_8. 
[8] Hong A, Zampieri RA, Shaw JJ, Floeter-Winter LM, Laranjeira-Silva MF. One [31] Van Der Spoel D, Lindahl E, Hess B, Groenhof G, Mark AE, Berendsen HJC. 
health approach to Leishmaniases: understanding the disease dynamics through GROMACS: fast, flexible, and free. J Comput Chem 2005;26:1701–18. https:// 
diagnostic tools. Pathogens 2020;9(2020):809–33. https://doi.org/10.3390/ doi.org/10.1002/JCC.20291. 
PATHOGENS9100809. Vol. 9, Page 809. [32] Abraham MJ, Murtola T, Schulz R, Páll S, Smith JC, Hess B, Lindah E. Gromacs: 
[9] Rogers ME, Bates PA. Leishmania manipulation of sand fly feeding behavior high performance molecular simulations through multi-level parallelism from 
results in enhanced transmission. PLoS Pathog 2007;3. https://doi.org/10.1371/ laptops to supercomputers. SoftwareX 2015;1:19–25. https://doi.org/10.1016/J. 
JOURNAL.PPAT.0030091. 0818–0825. SOFTX.2015.06.001. 
[10] Nes WD. Sterol methyl transferase: enzymology and inhibition. Biochim Biophys [33] Šudomová M, Hassan STS, Khan H, Rasekhian M, Nabavi SM. A multi- 
Acta, Mol Cell Biol Lipids 2000;1529:63–88. https://doi.org/10.1016/S1388- biochemical and in silico study on anti-enzymatic actions of pyroglutamic acid 
1981(00)00138-4. against PDE-5, ACE, and urease using various analytical techniques: unexplored 
[11] Mukherjee S, Xu W, Hsu FF, Patel J, Huang J, Zhang K. Sterol methyltransferase is pharmacological properties and cytotoxicity evaluation. Biomolecules 2019;9: 
required for optimal mitochondrial function and virulence in Leishmania major. 392–405. https://doi.org/10.3390/BIOM9090392. 
Mol Microbiol 2019;111:65–81. https://doi.org/10.1111/mmi.14139. [34] Trott O, Olson AJ, Vina AutoDock. Improving the speed and accuracy of docking 
[12] Urbina JA, Vivas J, Visbal G, Contreras LM. Modification of the sterol with a new scoring function, efficient optimization and multithreading. J Comput 
composition of Trypanosoma (Schizotrypanum) cruzi epimastigotes by Δ24(25)- Chem 2010;31:455–61. https://doi.org/10.1002/JCC.21334. 
sterol methyl transferase inhibitors and their combinations with ketoconazole. [35] Tian W, Chen C, Lei X, Zhao J, Liang J. CASTp 3.0: computed atlas of surface 
Mol Biochem Parasitol 1995;73:199 210. https://doi.org/10.1016/0166-6851 topography of proteins. Nucleic Acids Res 2018;46:W363. https://doi.org/ –
(95)00117-j. 10.1093/nar/gky473. –W367. 
[13] Reigada C, Sayé M, Valera-Vera E, Miranda MR, Pereira CA. Repurposing of [36] Kerwin SM. ChemBioOffice Ultra 2010 suite. J Am Chem Soc 2010;132:2466–7. 
terconazole as an anti Trypanosoma cruzi agent. Heliyon 2019;5:e01947/1- https://doi.org/10.1021/JA1005306. 
e01947/5. https://doi.org/10.1016/J.HELIYON.2019.E01947. 
14
P.O. Sakyi et al.                                                                                                                                                                                   I n  f o  r m   a t i c  s  i n   M   e d  i c i n  e   U  n  l o c  k  e d 37 (2023) 101162
[37] Sorokina M, Steinbeck C. Review on natural products databases: where to find [61] Ertl P, Schuffenhauer A. Estimation of synthetic accessibility score of drug-like 
data in 2020. J Cheminf 2020;12:1–51. https://doi.org/10.1186/S13321-020- molecules based on molecular complexity and fragment contributions. J Cheminf 
00424-9. 2009;1:1–11. https://doi.org/10.1186/1758-2946-1-8/TABLES/1. 
[38] Kwofie SK, Dolling NNO, Donkoh E, Laryea GM, Mosi L, Miller WA, [62] Speck-Planche A, Kleandrova VV, Luan F, Cordeiro M Natalia DS. Silico discovery 
Adinortey MB, Wilson MD. Pharmacophore-guided identification of natural and virtual screening of multi-target inhibitors for proteins in Mycobacterium 
products as potential inhibitors of mycobacterium ulcerans cystathionine tuberculosis. Comb Chem High Throughput Screen 2012;15:666–73. https://doi. 
γ-synthase metb. Computation 2021;9:1–24. https://doi.org/10.3390/ org/10.2174/138620712802650487. 
computation9030032. [63] Gupta Y, Maciorowski D, Zak SE, Jones KA, Kathayat RS, Azizi SA, Mathur R, 
[39] Mysinger MM, Carchia M, Irwin JJ, Shoichet BK. Directory of useful decoys, Pearce CM, Ilc DJ, Husein H, Herbert AS, Bharti A, Rathi B, Durvasula R, 
enhanced (DUD-E): better ligands and decoys for better benchmarking. J Med Becker DP, Dickinson BC, Dye JM, Kempaiah P. Bisindolylmaleimide IX: a novel 
Chem 2012;55:6582–94. https://doi.org/10.1021/jm300687e. anti-SARS-CoV2 agent targeting viral main protease 3CLpro demonstrated by 
[40] Goksuluk D, Korkmaz S, Zararsiz G, Karaagaoglu AE. easyROC: an interactive virtual screening pipeline and in-vitro validation assays. Methods 2021;195: 
web-tool for ROC curve analysis using R language environment. R J 2016;8: 57–71. https://doi.org/10.1016/j.ymeth.2021.01.003. 
213–30. [64] Sidhu KS, Bhangu SK, Pathak RK, Yadav IS, Chhuneja P. Identification of natural 
[41] Dallakyan S, Olson AJ. Small-molecule library screening by docking with PyRx. lead compounds for leaf rust of Wheat: a molecular docking and simulation study. 
Methods Mol Biol 2015;1263:243–50. https://doi.org/10.1007/978-1-4939- J Protein Proteonomics 2020;11:283–95. https://doi.org/10.1007/s42485-020- 
2269-7_19. 00048-5. 
[42] Daina A, Michielin O, Zoete V. SwissADME: a free web tool to evaluate [65] Agyapong O, Asiedu SO, Kwofie SK, Miller WA, Parry CS, Sowah RA, Wilson MD. 
pharmacokinetics, drug-likeness and medicinal chemistry friendliness of small Molecular modelling and de novo fragment-based design of potential inhibitors of 
molecules. Sci Rep 2017;7:42717/1-42717/13. https://doi.org/10.1038/ beta-tubulin gene of Necator americanus from natural products. Inform Med 
SREP42717. Unlocked 2021;26:100734/1-100734/20. https://doi.org/10.1016/j. 
[43] Sander T, Freyss J, Von Korff M, Rufener C. DataWarrior: an open-source program imu.2021.100734. 
for chemistry aware data visualization and analysis. J Chem Inf Model 2015;55: [66] Broni E, Kwofie SK, Asiedu SO, Miller WA, Wilson MD. A molecular modeling 
460–73. https://doi.org/10.1021/CI500588J. approach to identify potential antileishmanial compounds against the cell 
[44] Lagunin A, Stepanchikova A, Filimonov D, Poroikov V. PASS: prediction of division cycle (cdc)-2-Related kinase 12 (CRK12) receptor of leishmania 
activity spectra for biologically active substances. Bioinformatics 2000;16:747–8. donovani. Biomolecules 2021;11:458/1-458/32. https://doi.org/10.3390/ 
https://doi.org/10.1093/BIOINFORMATICS/16.8.747. biom11030458. 
[45] Kumari R, Kumar R, Lynn A. g_mmpbsa —a GROMACS tool for high-throughput [67] Pathania S, Singh PK, Narang RK, Rawal RK. Identifying novel putative ERK1/2 
MM-PBSA calculations. J Chem Inf Model 2014;54:1951–62. https://doi.org/ inhibitors via hybrid scaffold hopping –FBDD approach. J Biomol Struct Dyn 
10.1021/ci500020m. 2021;39:1–16. https://doi.org/10.1080/07391102.2021.1889670. 
[46] Kwofie SK, Kwarko GB, Broni E, Adinortey MB, Wilson MD. Molecular informatics [68] Dunn MF. Protein–ligand interactions: general description. ELS; 2010. p. 1–12. 
of trypanothione reductase of Leishmania major reveals novel chromen-2-one https://doi.org/10.1002/9780470015902.a0001340.pub2. 
analogues as potential leishmanicides. 2021, 100594/1-100594/36. https://doi. [69] Ikwu FA, Isyaku Y, Obadawo BS, Lawal HA, Ajibowu SA. In silico design and 
org/10.5772/INTECHOPEN.100594. IntechOpen. molecular docking study of CDK2 inhibitors with potent cytotoxic activity against 
[47] Dahiya R, Mohammad T, Gupta P, Haque A, Alajmi MF, Hussain A, Hassan MI. HCT116 colorectal cancer cell line. J Genet Eng Biotechnol 2020;18:1–12. 
Molecular interaction studies on ellagic acid for its anticancer potential targeting https://doi.org/10.1186/s43141-020-00066-2. 
pyruvate dehydrogenase kinase 3. RSC Adv 2019;9:23302–15. https://doi.org/ [70] Harel D, Khalid SA, Kaiser M, Brun R, Wünsch B, Schmidt TJ. Encecalol angelate, 
10.1039/C9RA02864A. an unstable chromene from Ageratum conyzoides L.: total synthesis and 
[48] Tippmann S. Programming tools: adventures with R. Nature 2014;517:109–10. investigation of its antiprotozoal activity. J Ethnopharmacol 2011;137:620–5. 
https://doi.org/10.1038/517109a. https://doi.org/10.1016/j.jep.2011.06.015. 
[49] Alkarkhi AFM, Alqaraghuli WAA. R statistical software. Appl Stat Environ Sci R [71] Arango V, Robledo S, Séon-Méniel B, Figadère B, Cardona W, Sáez J, Otálvaro F. 
2020;1:1–17. https://doi.org/10.1016/b978-0-12-818622-0.00002-2. Coumarins from Galipea panamensis and their activity against Leishmania 
[50] Du X, Li Y, Xia YL, Ai SM, Liang J, Sang P, Ji XL, Liu SQ. Insights into panamensis. J Nat Prod 2010;73:1012–4. https://doi.org/10.1021/NP100146Y. 
protein–ligand interactions: mechanisms, models, and methods. Int J Mol Sci [72] Kwofie S, Broni E, Yunus F, Nsoh J, Adoboe D, Miller W, Wilson M. Molecular 
2016;17:144/1-144/34. https://doi.org/10.3390/IJMS17020144. docking simulation studies identifies potential natural product derived- 
[51] Kandakatla N, Ramakrishnan G. Ligand based pharmacophore modeling and antiwolbachial compounds as filaricides against onchocerciasis. Biomedicines 
virtual screening studies to design novel HDAC2 inhibitors. Adv Bioinformatics 2021;9:1682/1-1682/33. https://doi.org/10.3390/biomedicines9111682. 
2014;2014:1–11. https://doi.org/10.1155/2014/812148. [73] Lipinski CA. Lead- and drug-like compounds: the rule-of-five revolution. Drug 
[52] John S, Thangapandian S, Arooj M, Hong JC, Kim KD, Lee KW. Development, Discov Today Technol 2004;1:337–41. https://doi.org/10.1016/J. 
evaluation and application of 3D QSAR Pharmacophore model in the discovery of DDTEC.2004.11.007. 
potential human renin inhibitors. BMC Bioinf 2011;12:1–14. https://doi.org/ [74] Kwofie SK, Annan DG, Adinortey CA, Boison D, Kwarko GB, Abban RA, 
10.1186/1471-2105-12-S14-S4/FIGURES/9. Adinortey MB. Identification of novel potential inhibitors of varicella-zoster virus 
[53] Andrade-Neto VV, Pereira TM, do Canto-Cavalheiro M, Torres-Santos EC. thymidine kinase from ethnopharmacologic relevant plants through an in-silico 
Imipramine alters the sterol profile in Leishmania amazonensis and increases its approach. J Biomol Struct Dyn 2021:1–17. https://doi.org/10.1080/ 
sensitivity to miconazole. Parasites Vectors 2016;91(9):1–8. https://doi.org/ 07391102.2021.1977700. 
10.1186/S13071-016-1467-8. 2016. [75] Ntie-Kang F, Zofou D, Babiaka SB, Meudom R, Scharfe M, Lifongo LL, Mbah JA, 
[54] Torres-Santos EC, Andrade-Neto VV, Cunha-Júnior EF, Do Canto-Cavalheiro MM, Mbaze LMa, Sippl W, Efange SMN. AfroDb: a select highly potent and diverse 
Atella GC, De Almeida Fernandes T, Costa PRR. Antileishmanial activity of natural product library from African medicinal plants. PLoS One 2013;8:e78085/ 
ezetimibe: inhibition of sterol biosynthesis, in vitro synergy with azoles, and 1-e78085/15. https://doi.org/10.1371/JOURNAL.PONE.0078085. 
efficacy in experimental cutaneous leishmaniasis. Antimicrob Agents Chemother [76] Kwofie SK, Broni E, Asiedu SO, Kwarko GB, Dankwa B, Enninful KS, Tiburu EK, 
2016;60:6844–52. https://doi.org/10.1128/AAC.01545-16. Wilson MD. Cheminformatics-based identification of potential novel anti-sars- 
[55] Empereur-Mot C, Guillemain H, Latouche A, Zagury JF, Viallon V, Montes M. cov-2 natural compounds of african origin. Molecules 2021;26:1–35. https://doi. 
Predictiveness curves in virtual screening. J Cheminf 2015;7:1–17. https://doi. org/10.3390/molecules26020406. 
org/10.1186/S13321-015-0100-8/TABLES/4. [77] Veber DF, Johnson SR, Cheng HY, Smith BR, Ward KW, Kopple KD. Molecular 
[56] Mandrekar JN. Receiver operating characteristic curve in diagnostic test properties that influence the oral bioavailability of drug candidates. J Med Chem 
assessment. J Thorac Oncol 2010;5:1315–6. https://doi.org/10.1097/ 2002;45:2615–23. https://doi.org/10.1021/JM020017N. 
JTO.0B013E3181EC173D. [78] van den Anker J, Reed MD, Allegaert K, Kearns GL. Developmental changes in 
[57] Hsin KY, Matsuoka Y, Asai Y, Kamiyoshi K, Watanabe T, Kawaoka Y, Kitano H. pharmacokinetics and pharmacodynamics. J Clin Pharmacol 2018;58:S10–25. 
systemsDock: a web server for network pharmacology-based prediction and https://doi.org/10.1002/JCPH.1284. 
analysis. Nucleic Acids Res 2016;44:W507–13. https://doi.org/10.1093/NAR/ [79] Sawale RT, Kalyankar TM, George R, Deosarkar SD. Molar refraction and 
GKW335. polarizability of antiemetic drug 4-amino-5-chloro-N-(2-(diethylamino)ethyl)-2 
[58] Tabrez S, Rahman F, Ali R, Muhammad F, Alshehri BM, Alaidarous MA, methoxybenzamide hydrochloride monohydrate in {Aqueous-Sodium or lithium 
Banawas S, Bin Dukhyil AA, Rub A. Repurposing of FDA-approved drugs as chloride} solutions at 30 o C. J Appl Pharmaceut Sci 2016;6:120–4. https://doi. 
inhibitors of sterol C-24 methyltransferase of Leishmania donovani to fight org/10.7324/JAPS.2016.60321. 
against leishmaniasis. Drug Dev Res 2021;82:1154–61. https://doi.org/10.1002/ [80] Flores-Sumoza M, Alcázar JJ, Márquez E, Mora JR, Lezama J, Puello E. Classical 
DDR.21820. QSAR and docking simulation of 4-pyridone derivatives for their antimalarial 
[59] Martín-Martínez M, Pérez-Gordillo FL, Álvarez de La Rosa D, Rodríguez Y, activity. Mol 2018;23:3166/1-3166/12. https://doi.org/10.3390/ 
Gerona-Navarro G, González-Muñiz R, Zhou MM. Modulating mineralocorticoid MOLECULES23123166. Page 3166. 23 (2018). 
receptor with non-steroidal antagonists. New opportunities for the development [81] Ertl P, Schuffenhauer A. Estimation of synthetic accessibility score of drug-like 
of potent and selective ligands without off-target side effects. J Med Chem 2017; molecules based on molecular complexity and fragment contributions. J Cheminf 
60:2629–50. https://doi.org/10.1021/ACS.JMEDCHEM.6B01065. 2009;1:1–11. https://doi.org/10.1186/1758-2946-1-8/TABLES/1. 
[60] Bramlage P, Swift SL, Thoenes M, Minguet J, Ferrero C, Schmieder RE. Non- [82] Alavijeh MS, Chishty M, Qaiser MZ, Palmer AM. Drug metabolism and 
steroidal mineralocorticoid receptor antagonism for the treatment of pharmacokinetics, the blood-brain barrier, and central nervous system drug 
cardiovascular and renal disease. Eur J Heart Fail 2016;18:28–37. https://doi. discovery. NeuroRx 2005;2:554–71. https://doi.org/10.1602/ 
org/10.1002/EJHF.444. NEURORX.2.4.554. 
15
P.O. Sakyi et al.                                                                                                                                                                                   I n  f o  r m   a t i c  s  i n   M   e d  i c i n  e   U  n  l o c  k  e d 37 (2023) 101162
[83] Glassman PM, Muzykantov VR. Pharmacokinetic and pharmacodynamic [105] Zhang G, Guo K, Wang P, Shan Y, Ma C. Identification of the new in vivo 
properties of drug delivery systems. J Pharmacol Exp Therapeut 2019;370: metabolites of ilaprazole in rat plasma after oral administration by LC-MS: in 
570–80. https://doi.org/10.1124/JPET.119.257113. silico prediction of the H+/K+-ATPase inhibitor. Molecules 2021;26:459–76. 
[84] Islam MA, Pillay TS. Identification of promising anti-DNA gyrase antibacterial https://doi.org/10.3390/MOLECULES26020459. 
compounds using de novo design, molecular docking and molecular dynamics [106] Gervazoni LFO, Barcellos GB, Ferreira-Paes T, Almeida-Amaral EE. Use of natural 
studies. J Biomol Struct Dyn 2019;38:1798–809. https://doi.org/10.1080/ products in leishmaniasis chemotherapy: an overview. Front Chem 2020;8:1031/ 
07391102.2019.1617785. 1-1031/43. https://doi.org/10.3389/FCHEM.2020.579891/BIBTEX. 
[85] Löbenberg R, Amidon GL, Ferraz HG, Bou-Chacra N. Mechanism of [107] Baloch N, Alkahraman YMSA, Zaidi MA, Madkour HMF. Evaluation of 6, 8- 
gastrointestinal drug absorption and application in therapeutic drug delivery. Dichloro-2-methyl-4H-Chromen-4-one derivatives as antileishmanial agents. Glob 
Ther Deliv Methods A Concise Overv Emerg Areas 2013:8–22. https://doi.org/ J Sci Front Res (GJSFR) 2012;12:2–8. 
10.4155/EBO.13.349/ASSET/IMAGES/LARGE/FIGURE6. JPEG. [108] Song C-H, Kim N, Kim D-H, Lee H-N, Surh Y-J. 17-β estradiol exerts anti- 
[86] Suenderhauf C, Hammann F, Huwyler J. Computational prediction of blood-brain inflammatory effects through activation of Nrf2 in mouse embryonic fibroblasts. 
barrier permeability using decision tree induction. Molecules 2012;17:10429–45. PLoS One 2019;14:e0221650/1-e0221650/13. https://doi.org/10.1371/journal. 
https://doi.org/10.3390/MOLECULES170910429. pone.0221650. 
[87] Broni E, Kwofie SK, Asiedu SO, Miller WA, Wilson MD. A molecular modeling [109] Wu X, Zhang S, Zhang Q, Zhao Y, Chen G, Guo W, Wang L. The contribution of 
approach to identify potential antileishmanial compounds against the cell specific subsites to catalytic activities in active site architecture of a GH11 
division cycle (Cdc)-2-related kinase 12 (crk12) receptor of leishmania donovani. xylanase. Appl Microbiol Biotechnol 2020;104:8735–45. https://doi.org/ 
Biomolecules 2021;11:458/1-458/32. https://doi.org/10.3390/biom11030458. 10.1007/S00253-020-10865-9. 
[88] Lin JH, Yamazaki M. Role of P-glycoprotein in pharmacokinetics: clinical [110] Azam SS, Abro A, Raza S, Saroosh A. Structure and dynamics studies of sterol 24- 
implications. Clin Pharmacokinet 2003;42:59–98. https://doi.org/10.2165/ C-methyltransferase with mechanism based inactivators for the disruption of 
00003088-200342010-00003. ergosterol biosynthesis. Mol Biol Rep 2014;41:4279–93. https://doi.org/ 
[89] Parasuraman S. Prediction of activity spectra for substances. J Pharmacol 10.1007/S11033-014-3299-Y. 
Pharmacother 2011;2:52–3. https://doi.org/10.4103/0976-500X.77119. [111] Asiedu SO, Kwofie SK, Broni E, Wilson MD. Computational identification of 
[90] Stepanchikova A, Lagunin A, Filimonov D, Poroikov V. Prediction of biological potential anti-inflammatory natural compounds targeting the p38 mitogen- 
activity spectra for substances: evaluation on the diverse sets of drug-like activated protein kinase (MAPK): implications for COVID-19-induced cytokine 
structures. Curr Med Chem 2012;10:225–33. https://doi.org/10.2174/ storm. Biomolecules 2021;11:653–78. https://doi.org/10.3390/BIOM11050653. 
0929867033368510. [112] Congreve M, Marshall F. The impact of GPCR structures on pharmacology and 
[91] Perez J, Fuertes M, Nguewa P, Castilla J, Alonso C. Anticancer compounds as structure-based drug design. Br J Pharmacol 2010;159:986–96. https://doi.org/ 
leishmanicidal drugs: challenges in chemotherapy and future perspectives. Curr 10.1111/J.1476-5381.2009.00476.X. 
Med Chem 2008;15:433–9. https://doi.org/10.2174/092986708783503221. [113] Grinter SZ, Zou X. Challenges, applications, and recent advances of protein-ligand 
[92] Sanderson L, Yardley V, Croft SL. Activity of anti-cancer protein kinase inhibitors docking in structure-based drug design. Molecules 2014;19:10150–76. https:// 
against Leishmania spp. J Antimicrob Chemother 2014;69:1888–91. https://doi. doi.org/10.3390/MOLECULES190710150. 
org/10.1093/JAC/DKU069. [114] Bronowska AK. Thermodynamics - interaction studies - Solids, liquids and gases. 
[93] Kalmi G, Vignon-Pennamen MD, Ram-Wolff C, Battistella M, Lafaurie M, Thermodyn. - Interact. Stud. - Solids, Liq. Gases 2011:1–49. https://doi.org/ 
Bouaziz JD, Hamane S, Bernard S, Bretagne S, Thiéblemont C, Bagot M, de 10.5772/823. 
Masson A. Visceral leishmaniasis in patients with lymphoma: case reports and [115] Estrada E, Uriarte E, Vilar S. Effect of protein backbone folding on the stability of 
review of the literature. Medicine (Baltimore) 2020;99:e22787/1-e22787/5. protein-ligand complexes. J Proteome Res 2006;5:105–11. https://doi.org/ 
https://doi.org/10.1097/MD.0000000000022787. 10.1021/pr0503174. 
[94] Machado PRL, Ribeiro CS, França-Costa J, Dourado MEF, Trinconi CT, [116] Kwofie S, Dankwa B, Enninful K, Adobor C, Broni E, Ntiamoah A, Wilson M. 
Yokoyama-Yasunaka JKU, Malta-Santos H, Borges VM, Carvalho EM, Uliana SRB. Molecular docking and dynamics simulation studies predict Munc18b as a target 
Tamoxifen and meglumine antimoniate combined therapy in cutaneous of mycolactone: a plausible mechanism for granule exocytosis impairment in 
leishmaniasis patients: a randomised trial. Trop Med Int Health 2018;23:936–42. buruli ulcer pathogenesis. Toxins (Basel) 2019;11:181/1-181/16. https://doi. 
https://doi.org/10.1111/TMI.13119. org/10.3390/toxins11030181. 
[95] Pulivarthi D, Steinberg KM, Monzote L, Piñón A, Setzer WN. Antileishmanial [117] Skjevik ÅA, Teigen K, Martinez A. Overview of computational methods employed 
activity of compounds isolated from sassafras albidum. Nat Prod Commun 2015; in early-stage drug discovery. Future Med Chem 2009;1:49–63. https://doi.org/ 
10:1229–30. https://doi.org/10.1177/1934578X1501000723. 10.4155/FMC.09.7. 
[96] Weidner S, Kittelmann M, Goeke K, Ghisalba O, Zähner H. 3’-Demethoxy-3’- [118] Mak KK, Pichika MR. Artificial intelligence in drug development: present status 
hydroxystaurosporine-O-methyltransferase from Streptomyces longisporoflavus and future prospects. Drug Discov Today 2019;24:773–80. https://doi.org/ 
catalyzing the last step in the biosynthesis of staurosporine. J Antibiot (Tokyo) 10.1016/J.DRUDIS.2018.11.014. 
1998;51:679–82. https://doi.org/10.7164/ANTIBIOTICS.51.679. [119] Bajorath J. Integration of virtual and high-throughput screening. Nat Rev Drug 
[97] Axelrod J, Daly J. Phenol-O-methyltransferase. BBA - Enzymol. 1968;159:472–8. Discov 2002;1:882–94. https://doi.org/10.1038/NRD941. 
https://doi.org/10.1016/0005-2744(68)90131-9. [120] Zheng M, Zhao J, Cui C, Fu Z, Li X, Liu X, Ding X, Tan X, Li F, Luo X, Chen K, 
[98] Coque JJR, Alvarez-Rodríguez ML, Larriba G. Characterization of an inducible Jiang H. Computational chemical biology and drug design: facilitating protein 
chlorophenol O-methyltransferase from Trichoderma longibrachiatum involved structure, function, and modulation studies. Med Res Rev 2018;38:914–50. 
in the formation of chloroanisoles and determination of its role in cork taint of https://doi.org/10.1002/MED.21483. 
wines. Appl Environ Microbiol 2003;69:5089–95. https://doi.org/10.1128/ [121] Riolo G, Cantara S, Marzocchi C, Ricci C. miRNA targets: from prediction tools to 
AEM.69.9.5089-5095.2003/ASSET/943B581A-42DB-4174-B2A8- experimental validation. Methods Protoc 2020;4:1–20. https://doi.org/10.3390/ 
9A9FE5E9C153/ASSETS/GRAPHIC/AM0930278003. JPEG. MPS4010001. 
[99] Chi Y, Gao K, Zhang H, Takeda M, Yao J. Suppression of cell membrane [122] Mateus A, Määttä TA, Savitski MM. Thermal proteome profiling: unbiased 
permeability by suramin: involvement of its inhibitory actions on connexin 43 assessment of protein state through heat-induced stability changes. Proteome Sci 
hemichannels. Br J Pharmacol 2014;171:3448–62. https://doi.org/10.1111/ 2017;15:e9232/1-e9232/11. https://doi.org/10.1186/S12953-017-0122-4. 
BPH.12693. [123] Das A, Das MC, Das N, Bhattacharjee S. Evaluation of the antileishmanial potency, 
[100] Sarkar P, Yarlagadda V, Ghosh C, Haldar J. A review on cell wall synthesis toxicity and phytochemical constituents of methanol bark extract of Sterculia 
inhibitors with an emphasis on glycopeptide antibiotics. Medchemcomm 2017;8: villosa. Pharm Biol 2017;55:998–1009. https://doi.org/10.1080/ 
516–33. https://doi.org/10.1039/C6MD00585C. 13880209.2017.1285946. 
[101] Poole K. Bacterial stress responses as determinants of antimicrobial resistance. [124] Silva J de C, de França PRL, Converti A, Porto TS. Kinetic and thermodynamic 
J Antimicrob Chemother 2012;67:2069–89. https://doi.org/10.1093/JAC/ characterization of a novel Aspergillus aculeatus URM4953 polygalacturonase. 
DKS196. Comparison of free and calcium alginate-immobilized enzyme. Process Biochem 
[102] Singh VS, Dhawale SC, Shakeel F, Faiyazuddin M, Alshehri S. Antiarthritic 2018;74:61–70. https://doi.org/10.1016/J.PROCBIO.2018.07.010. 
potential of calotropis procera leaf fractions in FCA-induced arthritic rats: [125] Baig MH, Ahmad K, Rabbani G, Danishuddin M, Choi I. Computer aided drug 
involvement of cellular inflammatory mediators and other biomarkers. design and its application to the development of potential drugs for 
Agriculture 2021;11:68–84. https://doi.org/10.3390/AGRICULTURE11010068. neurodegenerative disorders. Curr Neuropharmacol 2018;16:740–8. https://doi. 
[103] Hossen MA, Reza ASMA, Ahmed AMA, Islam MK, Jahan I, Hossain R, Khan MF, org/10.2174/1570159X15666171016163510. 
Maruf MRA, Haque MA, Rahman MA. Pretreatment of Blumea lacera leaves [126] Talele T, Khedkar S, Rigby A. Successful applications of computer aided drug 
ameliorate acute ulcer and oxidative stress in ethanol-induced Long-Evan rat: a discovery: moving drugs from concept to the clinic. Curr Top Med Chem 2010;10: 
combined experimental and chemico-biological interaction. Biomed 127–41. https://doi.org/10.2174/156802610790232251. 
Pharmacother 2021;135:111211/1-111211/11. https://doi.org/10.1016/J. [127] Ganeshpurkar A, Saluja AK. The pharmacological potential of rutin. Saudi 
BIOPHA.2020.111211. Pharmaceut J 2017;25:149–64. https://doi.org/10.1016/j.jsps.2016.04.025. 
[104] Adeola Falade V, Isaac Adelusi T, Olaide Adedotun I, Abdul-Hammed M, Alabi [128] Kant V, Ranjan R, Kumar P, Mandal D, Vijayakumar S. Silico screening , 
Lawal T, Alabi Agboluaje S, Silico I. Pharmacology, in silico investigation of molecular dynamic simulations , and in vitro activity of selected natural 
saponins and tannins as potential inhibitors of SARS-CoV-2 main protease (Mpro). compounds as an inhibitor of Leishmania donovani 3 - mercaptopyruvate 
Silico Pharmacol 2021;9:1–15. https://doi.org/10.1007/S40203-020-00071-W. sulfurtransferase. Parasitol Res 2022:2093. https://doi.org/10.1007/s00436-022- 
07532-5. –2109. 
16