QSAR of SARS-CoV-2 Main Protease Inhibitors Utilizing Theoretical Molecular Descriptors

doi:10.21203/rs.3.rs-1526870/v1

The COVID-19 caused by novel strain of severe acute respiratory syndrome coronaviruses (SARS-CoV-2), has caused fatalities around the world since the end of 2019 due to its high virulence and quick multiplicity in the human body. Hence there has been an exigency to develop effective remedial measures to mitigate the mortalities. Scientists have been able to develop few corona vaccines to provide immunity, but there are no specific small molecule chemotherapeutics to combat the novel coronavirus which has spread to the whole world due to its contagiousness. In the viral genome exploration, it has been found that the main protease also known as chymotrypsin-like cysteine protease ([M^pro] or 3C-like protease [3CL^pro]) is responsible for the novel corona virus replication, transcription and host immunity destruction. Therefore, the main protease has been selected as one of the major targets for the design of new inhibitors. The protein crystallographic and molecular docking studies on SARS-CoV-2 Mpro inhibitors and some QSAR studies have been carried out on SARS-CoV main protease inhibitors to get some lead molecules for SARS-CoV-2 inhibition. But there is hardly any QSAR done on the diverse data of SARS-CoV-2 main protease inhibitors. In view of it, QSAR studies have been attempted on SARS-CoV-2 M^pro inhibitors utilizing theoretical molecular descriptors solely computed from the structures of novel corona viral main protease inhibitors. The developed best QSAR model showing R², Q²_Looand R²_predvalues of 0.7389, 0.6666, and 0.6453 respectively has been further validated on external data set where a good correlation (r = 0.787) has been found. Therefore, this model may be useful for the design of new SARS-CoV-2 main protease inhibitors.

SARS-CoV-2

Mpro or 3CLpro main protease inhibitors

Computed structural descriptors

QSAR

Anti-COVID-19 drug design

The COVID-19 outbreak caused by a novel coronavirus strain (SARS-CoV-2) with both high fatality and potential for human transmission was first reported in December 2019 [1–2].The novel corona virus is too infectious to become pandemic within a very short span. Ithas infected more than 452 million lives in 227 countries and caused6 million deaths till March 10, 2022[3] and has also badly affected the social security and economy of the countries. The doctors prescribe the USFDA approved existing antibacterials (azithromycin, doxycycline), antimalarials (hydroxychloroquine, chloroquine), antivirals (ribavirin,favipiravir, remdesivir, nitazoxanide), anti-inflammatory (colchicine), estrogen agonist (raloxifene), vasodilator (nitric oxide), monoclonal antibody (mepolizumab), 2-deoxy-D-glucose (2-DG), anti-inflammatory immunosuppressant (methyl prednisolone), ACE2 receptor antagonist (losartan), ascorbic acid (vitamin c), and ambroxol which have been repurposed to combat the dreadful SARS-CoV-2 virulence [4]. As there is no other treatment option at an early stage of infection, the National Regulatory Authorities have granted emergency use of nineCOVID-19 vaccines viz. Covishield, Sputnik-V, Convideciam BIDMC, Covaxin, CoronaVac, BBIBP-CorV, Comirnaty andSpikeVax [5]. Although the vaccination drives are in place around the world, but covering the entire population in a short period is a challenging task for the countries having high population densities. The COVID-19 vaccines develop the host active immunity which is not retained for a long time. Therefore, even after vaccination the individuals get effected by COVID-19.The novel corona viral spike attaches to the host cell protein leading to fusion followed by invasion and replication [6–7].

The genomic components of SARS-CoV-2 are coded by chymotrypsin-like cysteine protease ([M^pro] or 3C-like protease [3CL^pro]) which helps in viral replication and transcription. These proteins are reported to inhibit host cell CDK4 and apoptosis, causing destruction of the host body immunity [8].The novel corona virus replication can be stopped by inhibiting the SARS-CoV-2 main protease. Therefore, SARS-CoV-2 main protease has been taken as one of the crucial targets for the anti COVID-19 drug design [9–10].

Structure based drug design utilizing the available crystal structure of SARS-CoV-2 main protease [11–12] and in silico molecular docking has been carried out to explore the molecular mechanisms of these drugs against SARS-CoV-2 main protease [13]. Further ligand based design of SARS-CoV-2 main protease inhibitors can be done utilizing QSAR modeling. Researchers have attempted QSAR modeling on SARS-CoV main protease inhibitors and have used those in HTS of lead inhibitors against COVID-19SARS-CoV-2 main protease [14–17].

Therefore, the main objective of the present work has been to develop statistically validated QSAR models utilizing computed structural descriptors of the SARS-CoV-2 main protease inhibitors and to test its predictability on the external data set. Such validated model may be useful in the identification new leads and design of SARS-CoV-2 main protease inhibitors.

Activity Data

A total number of 40 structurally diverse compounds (Table 1) having in vitro SARS-CoV-2 main protease inhibitory activity evaluated under the similar protocol has been taken from the literature [10, 18-25]. The SARS-CoV-2 main protease enzyme plays a key role in mediating viral replication and transcription. The enzyme inhibitory activities have been measured in terms of IC₅₀values measuring the affinity of the molecule to bind at the active site of SARS-CoV-2 main protease. The IC₅₀values were converted to pIC₅₀for QSAR analysis. The negative logarithm of these IC₅₀ values (pIC₅₀) has been taken as dependent variable whereas theoretical molecular descriptors have been computed and taken as independent variables for these inhibitors.

Methods of structural descriptor calculation

Structure computation and molecular optimization

The 2D chemical structures of 40 SARS-CoV-2 main protease inhibitors were drawn using Chemdraw 8.0 software [26]. The drawn 2D structures were converted into 3D files which were optimized using MM2 force field. The molecular optimization was carried out taking a convergence criterion and dielectric constant of 0.01 Kcal /mol. and 1.0 respectively [27] utilizing the Chem3D Ultra window.

Theoretical molecular descriptors computation

All the 3D mol files were browsed into PaDEL Descriptor Computation software for the calculation of theoretical molecular descriptors. The PaDEL is an open sourced freeware for the molecular property calculation [28]. A total number of 1875 one-dimensional, two-dimensional and three-dimensional descriptors were calculated followed by the pre-treatment of the descriptor data to reduce the descriptors to 539(Table 2). The descriptors with perfectly constant and highly inter-correlated descriptors were removed considering the variance and correlation coefficient cut-off values of 0.0001 and 0.85 using V-WSP algorithm [29] incorporated in the NanoBridges software [30]. This reduced descriptor data set has been taken into the QSAR modeling study.

Generation of QSAR models and statistical validation

As the number of structural descriptors is more than the number of observations (N), therefore, feature selection is necessary prior to the data division into training and test sets. Feature selection coupled with model development was done by genetic algorithm-multiple linear regression (GA-MLR) [31]. Many QSAR models were generated taking the combined descriptor data using GA-MLR methods [32-35] based on the theory of mutation and crossover of the parents’ genes to generate the new solutions taking the most appropriate transformations of the independent variables incorporated in the NanoBridges software [36]. A population of 100 different random combinations of the calculated molecular descriptors was generated taking default parameters as available in the NanoBridges software.

The effect of the computed descriptors on SARS-CoV-2 main protease inhibitory activities of these compounds has been shown through QSAR model development by taking each parent combination of descriptors for the entire data set using MLR. Before the QSAR model generation the data set was divided into test and training sets using the Kennard-Stone (KS) method [37].The models were validated using training and test set concept. The quality of model is denoted by R² (R is the square root of multiple R-square for regression), Q²_Loo(cross-validated R²) values for the training set, an external validation was performed by calculating predictive R² (R²_pred) and the standard error of estimation, SEE, represents standard deviation which is measured by the error mean square, which expresses the variation of the residuals or the variation about the regression line.

As per the criterion for a predictive and acceptable QSAR model, the values of Q²_Looand R²_pred should be more than 0.5 [38].The best training model was composed of 72.5% and the test set of 27.5% of the total data. The test compounds viz. 1, 7, 20, 24-28, 31, 35-37 (marked with^*) and the computed parameters used in QSAR model development for the total data are described in Table 3.

QSAR modeling

The effect of the individual and combination of 1D, 2D- and 3D-structural descriptors on the novel corona viral main protease inhibitory activities of the training set compounds were studied and the optimal QSAR model describing the variation of activity in terms of the descriptors is given as:

pIC50 = 0.08968+0.0004(+/-0.00018) AATSC7v +0.12985(+/-0.04054) maxHBint6–

0.00036(+/-0.00008)VR1_Dzi-0.23129(+/-0.13511)minsCH3Eq.1

N=28, R² = 0.6166, Q²_Loo = 0.4317_, R²pred = 0.5875, SEE = 0.4771

The model (Eq.1) describes an explained variance of 61.66% and an internal predicted variance of 43.17% of the observed data with the R²_pred,and SEE values as 0.5875, and 0.4771 respectively. The model has the poor statistical significance and lower Q²_Loo value which should be more than 0.5 as per the criterion for a predictive QSAR model. Therefore, outlier analyses were carried out by testing of applicability domain of the training QSAR model which determines its acceptance as per the OECD incorporated in NanoBridges software [39]. The compound nos. 21 and 20 of training and test sets respectively were detected as outliers. The removal of these compounds led to the development of the improved QSAR model, described by Eq.2 (Table 4).

pIC50 = 0.28522+0.00042(+/-0.00015) ATSC7v +0.15514(+/-0.03495) maxHBint10

0.00037(+/-0.00007) VR1_Dzi -0.35312(+/-0.11958) minsCH3…Eq.2

N=27, R²= 0.7389, Q²Loo = 0.6666, R²pred = 0.6453, SEE= 0.4009, F =15.5647 (DF :4, 22)

The developed QSAR model (2) explains the 73.89% variance of the SARS-CoV-2 main protease inhibitory activity of the deliberated compounds with good correlation coefficient value (R= 0.859) and Q2Loo value (0.6666) of variances. This model with good64.53% external predictability with SEE value of 0.4 is quite acceptable with high >99.9% of statistical significance (F_{4,22(α=0.001)}=6.81;F_4,22=15.5647). Hence this model also with the highly significant regression coefficient values for the four computed parameters described in table 4, was used to predict the test set compounds in the present study. The square correlation coefficient (R²=0.720) describing the 72% explained variance between the observed activities vs. predicted activities of the test set compounds with good correlation coefficient (R=0.848) value(Fig. 1) suggests good model predictivity.

According to the QSAR model 2, the parameters such as ATSC7v, and maxHBint10positively contributetowards SARS-CoV-2 main protease inhibition whereas the features such as VR1_Dzi and minsCH3have negative contribution to SARS-CoV-2 M^pro inhibition.

Validation of the developed model on external data set

In order to further check the validity of the above in silico developed four parametric computed descriptor-based QSAR model (Eq.2), the recently published in vitro external data set (Table 2; Sl. No. 41-79) available in published literature has been used where all the compounds have been tested in vitro against SARS-CoV-2 M^proin vitro cell biological screening almost similar to the compounds used in developing the model. The compounds 41-42, 43-52 and 53-79 have been taken from [40-42] respectively. In view of the wide diversity of the molecules in this dataset and limitation of the QSAR model predictability that the molecules to be predicted should fall within the parameters substituent spanned space (SSS), and should have similar structural scaffolds used in the training and test set used in developing the QSAR model, the molecules 46, 62, 69, and 70having different structural scaffolds and the molecules 64, 65, 71, 73, 75, 76 and 77 with computed parameters values beyond the limits used in developing the QSAR model were excluded from the external dataset. The activity of the rest of the 28 compounds was predicted from the QSAR model 2. The calculated modeled parameters, the observed and predicted activities (pIC₅₀ in µM) are described in Table 5.

The squared correlation (R²=0.620) between the observed versus predicted activities describing the explained variance with a reasonably good correlation (R= 0.7874) (Figure 2) further validates the developed model 2.

Despite the advances achieved in the vaccine research and new vaccines development, there is a pressing need to design and develop specific small molecule that can inhibit the genomic constituents of SARS-CoV-2 main protease responsible for the viral replication and host immunity destruction. Thoughthe structure based drug design on several molecules in this paper are reported but hardly any attempt has been made to develop QSAR models on these novel corona virus main protease inhibitors belonging to diverse chemical prototypes. Hence the QSAR model developed utilizing theoretical molecular descriptors are the maiden attempt in this direction to the best of our knowledge. The best developed model(Eq. 2) is statistically validated on internal and external data sets.The positive contribution of ATSC7v and maxHBint10 in this model shows that increase in vander Waals and hydrogen bond interactions may increase the enzyme inhibition while the increase in ionizing potential (VR1_Dzi) and electrotopological state atom type (minsCH3) featuresmay have detrimental effect towards the inhibition of SARS-CoV-2 main protease. The good predictability of the developed QSAR model on the external data set of 28 diverse molecules as shown by good correlation (R = 0.787) and the explained variance (62%) suggest that this model may be useful in accelerating the discovery of newer, cost effective and promising SARS-CoV-2 main protease inhibitors to combat the dreadedCOVID-19 in the future.

Acknowledgements

Authors acknowledge the authority of the corresponding Institutes for the support. Authors thank Mr. Sarfaraz Ahmed and Ankit Kumar Gupta for their assistance.

Funding

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Conflicts of interest/Competing interests

Authors declare that there is no conflict of interest.

Availability of data and material

Not applicable

Code availability

Not applicable

Authors' contributions

SN: The manuscript write up, data generation, model interpretation, and discussion; MK: model

development; AKS: supervise the project.

Zhou P, Yang XL, Wang XG, Hu B, Zhang L, Zhang W, Si HR, Zhu Y, Li B, Huang CL, Chen HD, Chen J, Luo Y, Guo H, Jiang RD, Liu MQ, Chen Y, Shen XR, Wang X, Zheng XS, Zhao K, Chen QJ, Deng F, Liu LL, Yan B, Zhan FX, Wang YY, Xiao GF Shi ZL(2020)A pneumonia outbreak associated with a new coronavirus of probable bat origin.Nature579:270–273
Wu F, Zhao S, Yu B, Chen YM, Wang W, Song ZG, Hu Y, Tao ZW, Tian JH, Pei YY, Yuan ML, Zhang YL, Dai FH, Liu Y, Wang QM, Zheng JJ, Xu L, Holmes EC, Zhang YZ (2020) A new coronavirus associated with human respiratory disease in China, vol 579. Nature, pp 265–269
https://wwwworldometersinfo/coronavirus/(accessed on March 10 (2022) )
Chen PL, Lee NY, Cia CT, Ko WC, Hsueh PR (2020) A Review of Treatment of Coronavirus Disease 2019 (COVID-19): Therapeutic Repurposing and Unmet Clinical Needs. Front Pharmacol 1:584956
https://wwwwhoint/emergencies/diseases/novel-coronavirus-2019/covid-19-vaccines (accessed on March 10, 2022)
Belouzard S, Millet JK, Licitra BN, Whittaker GR (2012) Mechanisms of coronavirus cell entry mediated by the viral spike protein. Viruses 4:1011–1033
Lan J, Ge J, Yu J, Shan S, Zhou H, Fan S, Zhang Q, Shi X, Wang Q, Zhang L, Wang X (2020) Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor. Nature 581:215–220
Báez-Santos YM, St John SE, Mesecar AD (2015) The SARS-coronavirus papain-like protease: structure, function and inhibition by designed antiviral compounds Antiviral. Res 115:21–38
Ullrich S, Nitsche C (2020) The SARS-CoV-2 main protease as drug target. Bioorg Med Chem Lett 30:127377
Jin Z, Du X, Xu Y, Deng Y, Liu M, Zhao Y, Zhang B, Li X, Zhang L, Peng C, Duan Y, Yu J, Wang L, Yang K, Liu F, Jiang R, Yang X, You T, Liu X, Yang X, Bai F, Liu H, Liu X, Guddat LW, Xu W, Xiao G, Qin C, Shi Z, Jiang H, Rao Z, Yang H (2020) Structure of Mpro from SARS-CoV-2 and discovery of its inhibitors. Nature 582:289–293
Douangamath A, Fearon D, Gehrtz P, Krojer T, Lukacik P, Owen CD, Resnick E, Strain-Damerell C, Aimon A, Ábrányi-Balogh P, Brandão-Neto J, Carbery A, Davison G, Dias A, Downes TD, Dunnett L, Fairhead M, Firth JD, Jones SP, Keeley A, Keserü GM, Klein HF, Martin MP, Noble MEM, O'Brien P, Powell A, Reddi RN, Skyner R, Snee M, Waring MJ, Wild C, London N, von Delft F, Walsh MA (2020) Crystallographic and electrophilic fragment screening of the SARS-CoV-2 main protease. Nat Commun 11:5047
Noske GD, Nakamura AM, Gawriljuk VO, Fernandes RS, Lima GMA, Rosa HVD, Pereira HD, Zeri ACM, Nascimento AFZ, Freire MCLC, Fearon D, Douangamath A, von Delft F, Oliva G, Godoy AS (2021) A Crystallographic Snapshot of SARS-CoV-2 Main Protease Maturation Process. J Mol Biol 433:167118
Nandi S, Kumar M, Saxena M, Saxena AK (2021) The Antiviral and Antimalarial Drug Repurposing in Quest of Chemotherapeutics to Combat COVID-19 Utilizing Structure-Based Molecular Docking. Comb Chem High Throughput Screen 24:1055–1068
Alves VM, Bobrowski T, Melo-Filho CC, Korn D, Auerbach S, Schmitt C, Muratov EN, Tropsha A (2021) QSAR Modeling of SARS-CoVMpro Inhibitors Identifies Sufugolix, Cenicriviroc, Proglumetacin, and other Drugs as Candidates for Repurposing against SARS-CoV-2. Mol Inf 40:e2000113
Kumar V, Roy K (2020) Development of a simple, interpretable and easily transferable QSAR model for quick screening antiviral databases in search of novel 3C-like protease (3CLPro enzyme inhibitors against SARS-CoV diseases SAR. QSAR Environ Res 31:511–526
Masand VH, Akasapu S, Gandhi A, Rastija V, Patil MK (2020) Structure features of peptide-type SARS-CoV main protease inhibitors: Quantitative structure activity relationship study. Chemometr Intell Lab Syst 5:206104172
Masand VH, Rastija V, Patil MK, Gandhi A, Chapolikar A (2020) Extending the identification of structural features responsible for anti-SARS-CoV activity of peptide-type compounds using QSAR modelling. SAR QSAR Environ Res 31:643–654
Dai W, Zhang B, Jiang XM, Su H, Li J, Zhao Y, Xie X, Jin Z, Peng J, Liu F, Li C, Li Y, Bai F, Wang H, Cheng X, Cen X, Hu S, Yang X, Wang J, Liu X, Xiao G, Jiang H, Rao Z, Zhang LK, Xu Y, Yang H, Liu H (2020) Structure-based design of antiviral drug candidates targeting the SARS-CoV-2 main protease. Science 368:1331–1335
Zhang L, Lin D, Sun X, Curth U, Drosten C, Sauerhering L, Becker S, Rox K, Hilgenfeld R (2020) Crystal structure of SARS-CoV-2 main protease provides a basis for design of improved α-ketoamide inhibitors. Science 368:409–412
Ma C, Sacco MD, Hurst B, Townsend JA, Hu Y, Szeto T, Zhang X, Tarbet B, Marty MT, Chen Y, Wang J (2020) Boceprevir, GC-376, and calpain inhibitors II, XII inhibit SARS-CoV-2 viral replication by targeting the viral main protease. BioRxiv doi:101101/20200420051581
Su H, Yao S, Zhao W, Li M, Liu J, Shang W, Xie H, Ke C, Gao M, Yu K, Liu H, Shen J, Tang W, Zhang L, Zuo J, Jiang H, Bai F, Wu Y, Ye Y, Xu Y (2020) Discovery of baicalin and baicalein as novel, natural product inhibitors of SARS-CoV-23CL protease in vitro. BioRxiv. doi:101101/20200413038687
Li G, Sun J, Li Y, Shi Y, Zhao J, Zhang TY, Zhang X (2020) Enantiomers of Chloroquine and Hydroxychloroquine Exhibit Different Activities Against SARS-CoV-2 in vitro, Evidencing S-Hydroxychloroquine as a Potentially Superior Drug for COVID-19. bioRxiv. doi:101101/20200526114033
Caly L, Druce JD, Catton MG, Jans DA, Wagstaff KM (2020) The FDA-approved drug ivermectin inhibits the replication of SARS-CoV-2 in vitro. Antiviral Res 178:104787
Sacco MD, Ma C, Lagarias P, Gao A, Townsend JA, Meng X, Dube P, Zhang X, Hu Y, Kitamura N, Hurst B, Tarbet B, Marty MT, Kolocouris A, Xiang Y, Chen Y, Wang J (2020) Structure and inhibition of the SARS-CoV-2 main protease reveal strategy for developing dual inhibitors against Mpro and cathepsin L. Sci Adv 6:eabe0751
Ghahremanpour MM, Tirado-Rives J, Deshmukh M, Ippolito JA, Zhang CH, Cabeza de Vaca I, Liosi ME, Anderson KS, Jorgensen WL (2020) Identification of 14 Known Drugs as Inhibitors of the Main Protease of SARS-CoV-2. ACS Med Chem Lett 11:2526–2533
Mills N (2006) ChemDraw Ultra 100. J Am Chem Soc 128:13649–13650
Halgren TA (1996) Merck molecular force field III Molecular geometries and vibrational frequencies. J Comput Chem 17:553–586
Yap CW (2011) PaDEL-descriptor: An open-source software to calculate molecular descriptors and fingerprints. J Comput Chem 32:1466–1474
Ballabio D, Consonni V, Mauri A, Claeys-Bruno M, Sergent M, Todeschini R (2014) A novel variable reduction method adapted from space-filling designs. Chemometr Intell Lab Syst 136:147–154
Ambure AP, Gajewicz RB, Puzyn AT, Roy K (2015) “NanoBRIDGES” software, Open access tools to perform QSAR and nano-QSAR modeling.ChemometrIntell. Lab Syst 147:1–13
Broadhurst D, Goodacre R, Jones A, Rowland JJ, Kell DB (1997) Genetic algorithms as a method for variable selection in multiple linear regression and partial least squares regression with applications to pyrolysis mass spectrometry. Anal Chim Acta 348:71–86
Hoffman BT, Kopajtic T, Katz JL, Newman AH (2000) 2D QSAR modelling and preliminary database searching for dopamine transporter inhibitors using genetic algorithm variable selection of Molconn Z descriptors. J Med Chem 43:4151–4159
Saxena M, Nandi S, Saxena AK (2019) QSAR and molecular docking studies of lethal factor protease inhibitors against Bacillus anthracis. SAR QSAR Environ Res 30:715–731
Nandi S, Ahmed S, Saxena AK (2018) Combinatorial design and virtual screening of potent anti-tubercular fluoroquinolone and isothiazoloquinolone compounds utilizing QSAR and pharmacophore modelling. SAR QSAR Environ Res 29:151–170
Saxena AK, Prathipati P (2003) Comparison of MLR, PLS and GA-MLR in QSAR analysis. SAR QSAR Environ Res 14:433–445
de Campos LJ, de Melo EB (2014) Modeling structure-activity relationships of prodiginines with antimalarial activity using GA/MLR and OPS/PLS. J Mol Graph Model 54:19–31
Kennard RW, StoneLA (1969) Computer aided design of experiments. Technometric 11:137–148
Golbraikh A, Tropsha A (2000) Beware of q2! J Mol Graph Model 20:269–276
Jaworska J, Nikolova-Jeliazkova N, Aldenberg T (2005) QSAR applicabilty domain estimation by projection of the training set descriptor space: a review. Altern Lab Anim 33:445–459
Vuong W, Khan MB, Fischer C, Arutyunova E, Lamer T, Shields J, Saffran HA, McKay RT, van Belkum MJ, Joyce MA, Young HS, Tyrrell DL, Vederas JC, Lemieux MJ (2020) Feline coronavirus drug inhibits the main protease of SARS-CoV-2 and blocks virus replication. Nat Commun 11:4282
Rathnayake AD, Zheng J, Kim Y, Perera KD, Mackin S, Meyerholz DK, Kashipathy MM, Battaile KP, Lovell S, Perlman S, Groutas WC, Chang KO (2020) 3C-like protease inhibitors block coronavirus replication in vitro and improve survival in MERS-CoV-infected mice. Sci Transl Med 12:eabc5332
Zhu W, Xu M, Chen CZ, Guo H, Shen M, Hu X, Shinn P, Klumpp-Thomas C, Michael SG, Zheng W (2020) Identification of SARS-CoV-2 3CL protease inhibitors by aquantitative high-throughput screening. ACS Pharmacol Transl Sci 3:1008–1016

Tables 1 and 5 are available in the Supplementary Files section.

Table 2. Descriptors used in current study.

AlogP, ALogp2, AMR, naAromAtom, nAtom, nN, nO, ATS0m, ATS3m, ATS0s, AATS0m, AATS1m, AATS5m, AATS6m, AATS7m, AATS8m, AATS0v, AATS3v, AATS5v, AATS8v, AATS0e, AATS3e, AATS5e, AATS7e, AATS8e, AATS0p, AATS5p, AATS7p, AATS8p, AATS1i, AATS2i, AATS3i, AATS4i, AATS5i, AATS6i, AATS7i, AATS8i, AATS1s, AATS6s, ATSC0c, ATSC3c, ATSC4c, ATSC5c, ATSC6c, ATSC7c, ATSC8c, ATSC0m, ATSC1m, ATSC2m, ATSC3m, ATSC4m, ATSC5m, ATSC6m, ATSC7m, ATSC8m, ATSC1v, ATSC2v, ATSC3v, ATSC4v, ATSC5v, ATSC6v, ATSC7v, ATSC8v, ATSC1e, ATSC2e, ATSC3e, ATSC4e, ATSC5e, ATSC6e, ATSC7e, ATSC8e, ATSC0p, ATSC1p, ATSC2p, ATSC4p, ATSC5p, ATSC6p, ATSC7p, ATSC8p, ATSC1i, ATSC2i, ATSC3i, ATSC4i, ATSC5i, ATSC6i, ATSC7i, ATSC8i, ATSC1s, ATSC2s, ATSC3s, ATSC5s, ATSC7s, AATSC1m, AATSC2m, AATSC3m, AATSC4m, AATSC5m, AATSC6m, AATSC7m, AATSC8m, AATSC0v, AATSC1v, AATSC3v, AATSC4v, AATSC5v, AATSC6v, AATSC7v, AATSC8v, AATSC2e, AATSC4e, AATSC5e, AATSC8e, AATSC0p, AATSC1p, AATSC2p, AATSC4p, AATSC5p, AATSC6p, AATSC7p, AATSC0i, AATSC1i, AATSC3i, ATSC4i, AATSC5i, AATSC6i, AATSC7i, AATSC8i, AATSC0s, AATSC1s, AATSC3s, AATSC6s, AATSC7s, MATS1c, MATS2c, MATS3c, MATS4c, MATS5c, MATS6c, MATS7c, MATS8c, MATS6m, MATS7m, MATS8m, MATS1e, MATS2e, MATS3e, MATS6e, MATS7e, MATS8e, MATS7p, MATS1s, MATS2s, MATS3s, MATS4s, MATS5s, MATS6s, MATS7s, MATS8s, GATS1c, GATS2c, GATS3c, GATS4c, GATS5Cm, GATS6c, GATS7c, GATS8c, GATS1m, GATS2m, GATS3m, GATS4m, GATS5m, GATS6m, GATS7m, GATS8m, GATS2v, GATS6v, GATS7v, GATS8v, GATS1e, GATS2e, GATS3e, GATS5e, GATS6e, GATS7e, GATS8e, GATS1p , GATS3p, GATS6p, GATS7p, GATS8p, GATS1i, GATS2i, GATS4i, GATS5i, GATS6i, GATS7i, GATS8i, GATS1s, GATS3s, GATS4s, GATS5s, GATS7s, GATS8s, SpMAD_DzZ, SM1_DzZ, VE1_DzZ, VE3_DzZ, VR1_DzZ, VR2_DzZ, VE1_Dzv, VE3_Dzv, VR1_Dzv, VE1_Dzp, VR1_Dzp, VE3_Dzi, VR1_Dzi, SpMAD_Dzs, SM1_Dzs, VE1_Dzs, VE3_Dzs, VR1_Dzs, VR2_Dzs, nBase, BCUTw-1l, BCUTw-1h, BCUTp-1l, BCUTp-1h, nBondsD, nBondsD2, nBondsM, SpMax1_Bhm, SpMax2_Bhm, SpMax3_Bhm, SpMax4_Bhm, SpMax7_Bhm, SpMin1_Bhm, SpMin2_Bhm, SpMin3_Bhm, SpMin4_Bhm, SpMin5_Bhm, SpMax1_Bhv, SpMax2_Bhv, SpMax3_Bhv, SpMin2_Bhe, SpMin1_Bhp, SpMin2_Bhp, SpMax1_Bhs, SpMax2_Bhs, SpMax4_Bhs, SpMax5_Bhs, SpMin1_Bhs, SpMin2_Bhs, SpMin5_Bhs, C1SP2, C2SP2, C3SP2, C1SP3, C2SP3, C3SP3, SCH-6, SCH-7, SC-3, SC-4, SC-5, VC-3, VPC-5, SP-5, ASP-0, ASP-1, ASP-2, VP-5, AVP-0, AVP-2, AVP-3, CrippenLogP, SpMAD_Dt, VE1_Dt, VE3_Dt, VR1_Dt, nHBd, nHBint2, nHBint3, nHBint4, nHBint5, nHBint6, nHBint7, nHBint8, nHBint9, nHBint10, nHssNH, nHdsCH, nHCsats, nHCsatu, nsCH3, nssCH2, nsssCH, naasC, nsssN, ndO, nssO, SHBd, SwHBa, SHBint3, SHBint4, SHBint5, SHBint6, SHBint7, SHBint8, SHBint9, SHBint10, SssCH2, SsssCH, SdssC, SaasC, minHBd, minHBa, minwHBa, minHBint2, minHBint3, minHBint4, minHBint5, minHBint6, minHBint7, minHBint8, minHBint9, minHBint10, minHssNH, minHdsCH, minHaaCH, minHCsats, minHCsatu, minHother, minsCH3, minssCH2, mindsCH, minaaCH, minsssCH, mindssC, minaasC, minssssC, minssNH, minsssN, minsOH, mindO, minssO, maxHBd, maxHBa, maxwHBa, maxHBint2, maxHBint3, maxHBint4, maxHBint5, maxHBint6, maxHBint7, maxHBint8, maxHBint9, maxHBint10, maxHsOH, maxHCsats, maxHother, maxssCH2, maxdsCH, maxsssCH, maxdssC, maxaasC, hmax, hmin, gmin, LipoaffinityIndex, ETA_dAlpha_B, ETA_Epsilon_4, ETA_Epsilon_5, ETA_dEpsilon_B, ETA_dEpsilon_D, ETA_Shape_Y, ETA_Beta, ETA_BetaP_s, ETA_dBetaP, ETA_Beta_ns_d, ETA_EtaP_F, FMF, nHBAcc, IC0, IC1, IC2, IC4, SIC0, SIC1, SIC2, SIC3, CIC1, CIC2, MIC0, MIC1, MIC2, MIC5, ZMIC1, ZMIC2, ZMIC3, Kier3, nAtomLC, nAtomP, nAtomLAC, MlogP, MDEC-11, MDEC-12, MDEC-22, MDEC-23, MDEC-24, MDEC-33, MDEC-34, MDEO-11, MDEO-12, MDEN-22, MDEN-23, MLFER_A, MLFER_BH, MLFER_S, MLFER_E, MLFER_L, piPC3, piPC5, R_TpiPCTPC, PetitjeanNumber, nRing, n5Ring, n6Ring, nG12Ring, nFRing, nF9Ring, nF10Ring, nTRing, nTG12Ring, nHeteroRing, n6HeteroRing, nF10HeteroRing, nRotB, nRotBt, RotBtFrac, LipinskiFailures, JGI1, JGI2, VE1_D, VE3_D, VR1_D, TopoPSA, SRW5, XlogP, TDB5u, TDB7u, TDB10u, TDB9m, TDB10m, TDB6v, TDB7v, TDB8v, TDB9v, TDB10v, TDB1e, TDB5e, TDB7e, TDB9e, TDB8p, TDB9p, TDB1i, TDB2i, TDB3i, TDB4i, TDB5i, TDB6i, TDB8s, TDB10s, PPSA-1, PPSA-2, PNSA-1, PNSA-2, PNSA-3, DPSA-1, FPSA-1, FPSA-2, WNSA-1, WNSA-2, RNCG, RNCS, THSA, TPSA, RHSA, LOBMAX, MOMI-X, MOMI-Z, MOMI-XY, MOMI-XZ, geomRadius, geomShape, RDF20u, RDF105u, RDF135u, RDF155u, RDF20m, RDF30m, RDF35m, RDF40m, RDF45m, RDF50m, RDF70m, RDF80m, RDF90m, RDF95m, RDF100m, RDF110m, RDF130m, RDF145m, RDF150m, RDF155m, RDF20p, RDF25s, RDF30s, RDF60s, L2u, L3u, P1u, P2u, E1u, E2u, E3u, Du, L1m, E1m, E2m, E3m, Vm, Dm, E1v, E2v, E3v, Dv, E3e, E3p, L1i.

Table 3. The modeled descriptor values for all 40 compounds.

Comp. no.	ATSC7v	maxHBint10	VR1_Dzi	minsCH3
^*1	-395.458	0	130.1922	0
2	-734.29	0	79.38738	2.327001
3	-372.586	0	590.5438	0
4	37.10563	0	150.7602	2.070666
5	-397.714	0	241.5008	1.85527
6	210.0141	5.246226	479.0142	0
^*7	383.3142	5.39812	395.2159	0
8	-192.603	5.059775	461.4965	0
9	-1255.7	6.28981	525.3916	1.70658
10	-1003.99	6.592604	609.8805	1.695154
11	-1035.51	0	776.8339	1.635225
12	-794.517	3.116049	671.0153	1.851533
13	-832.848	0.61348	704.2953	1.95825
14	-413.44	0	527.3441	1.94064
15	71.17779	0	276.89	1.963841
16	-679.212	0	373.8613	1.362426
17	-57.2659	0	325.0109	1.989577
18	-558.278	0	391.6459	1.367466
19	-122.1	5.244607	476.7568	1.887279
^*20	96.45326	0	589.0031	1.447895
21	-286.661	5.60415	987.2066	0
22	-251.624	0	226.6374	0
23	-643.549	0	245.465	2.235752
^*24	-643.549	0	244.782	2.235752
^*25	-643.549	0	245.465	2.235752
^*26	-518.413	0	189.0643	2.14464
^*27	-518.413	0	188.9529	2.14464
^*28	-518.413	0	189.0643	2.14464
29	-935.213	2.68039	848.2537	1.478692
30	997.5656	2.215466	2125.822	1.573871
^*31	-1214.86	5.031532	547.9355	1.483858
32	-924.857	5.224689	6212.701	1.472423
33	-1379.01	4.966731	590.2184	1.724453
34	-245.59	0	475.1876	1.368338
^*35	56.70054	4.923686	383.4664	1.912241
^*36	469.1809	5.192444	472.2942	0
^*37	-218.863	5.190335	1232.807	1.894632
38	-562.115	0.865989	488.4981	1.261853
39	-1561.84	0	1660.779	1.259306
40	-763.465	0	460.8289	1.668986

*test set

Table 4:Physical interpretation of the descriptor sused in QSAR model equation 2

Parameters	Physical interpretation
ATSC7v	Centred Broto-Moreau autocorrelation of lag 7 weighted by van der Waals volume
maxHBint10	Maximum E-State descriptors of strength for potential Hydrogen Bonds of path length 10
VR1_Dzi	Randic-like eigenvector-based index from Barysz matrix weighted by ionization potential
minsCH3	Minimum atom-type E-State: -CH3

Table1and5.docx

QSAR of SARS-CoV-2 Main Protease Inhibitors Utilizing Theoretical Molecular Descriptors

Status:

Version 1

Abstract

Figures

Introduction

Experimental Methods

Results And Discussion

Conclusion

Declarations

References

Table 1 and 5

Tables 2-4

Supplementary Files

Status:

Version 1