Submit to IJMS Review for IJMS Propose a Special Issue

Journal Menu

Journal Browser

Chemoinformatics and Bioinformatics Tools in Structure-Activity Modelling in Molecular Sciences

Special Issue Editors
Special Issue Information
Keywords
Published Papers

A special issue of International Journal of Molecular Sciences (ISSN 1422-0067). This special issue belongs to the section "Molecular Informatics".

Deadline for manuscript submissions: closed (31 March 2022) | Viewed by 29308

Share This Special Issue

Special Issue Editor

Dr. Bono Lučić

E-Mail Website
Guest Editor

NMR Centre, The Ruđer Bošković Institute, 10000 Zagreb, Croatia
Interests: chemoinformatics; structural bioinformatics; structure–activity modeling; QSAR; QSPR; molecular modeling; computational chemistry; molecular structural biophysics; development of model validation algorithms; variable selection algorithms; classification modeling; chance accuracy estimation; development of accuracy parameters; computational research in bioprospecting research; protein structure analysis and prediction
Special Issues, Collections and Topics in MDPI journals

Special Issue Information

Dear colleagues,

At a time of universal digitization of data in various fields of research, including molecular sciences, there are more and more studies modeling continuous or classification endpoints (activities/properties) of molecules. In doing so, endpoints of molecules are most often classified (digitized) into two classes—active or inactive, and the classification is often carried out by grouping data into three or more classes.

Quantitative structure–activity/property relationships (QSAR/QSPR) are the most common, but not the only, forms of structure–endpoint models in molecular sciences. The accuracy of models is expressed by validation procedures, and many quality parameters are defined in the OECD document related to regulatory structure–activity models for the purpose of health and environmental protection [1]. In this document, the accuracy parameters of classification models are very sparsely presented. However, numerous accuracy parameters are used today, and those used for classification models are calculated from the confusion matrix elements [2]. There is also an increased need for a better definition of procedures for validation of regulatory structure–activity models in the OECD document [1]. Their application in environmental and health protection (toxicity, bioavailability, sorption, biodegradability, etc.) has been defined by EU REACH regulations [3].

The development of structure–activity modeling of different types of endpoints of molecules (usually various types of biological activities) is accelerated using chemoinformatics and bioinformatics tools, servers, algorithms, and databases developed for small molecules and proteins.

The research activities in the development of novel chemoinformatics and bioinformatics tools are particularly important topics for this Special Issue, such as the development of:

- Valuable databases, servers, and data mining tools;

- Drug or lead structure identification or dereplication approaches used in bioprospecting research;

- Structure optimization tools;

- Molecular descriptors;

- Modeling and variable selection algorithms;

- Computational model validation methods;

- Multivariate linear and nonlinear methods;

- Machine learning and deep learning algorithms;

- Predictive or descriptive structure–activity models;

- Different visualization tools;

- Protein–ligand (target/small compound) interactions;

- Molecular docking, etc.

All these topics are of the highest importance for structure–activity modeling in molecular sciences.

This Special Issue aims to collect relevant contributions (papers) belonging to one or more of the topics listed above (and those related to them), which are important for the acceleration of structure–activity research in molecular sciences. Applications aimed at modeling a broad spectrum of chemical, biological, pharmaceutical, biochemical, and environmentally relevant activities and properties of molecules are also appreciated.

All forms of scientific articles covering mentioned or related topics are welcomed, i.e., original papers, reviews, and communications.

[1] Guidance Document on the Validation of (Quantitative) Structure-Activity Relationship [(Q)SAR] Models, [https://www.oecd.org/env/guidance-document-on-the-validation-of-quantitative-structure-activity-relationship-q-sar-models-9789264085442-en.htm]

[2] D. M. W. Powers, Evaluation: from precision, recall and f-measure to roc, informedness, markedness & correlation, J. Machine Learning Techn., 2011, 2, 37-63

[3] Regulation (EC) No 1907/2006: REACH - Registration, Evaluation, Authorisation and Restriction of Chemicals. [http://ec.europa.eu/enterprise/sectors/chemicals/reach/index_en.htm].

Dr. Bono Lučić
Guest Editor

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. International Journal of Molecular Sciences is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. There is an Article Processing Charge (APC) for publication in this open access journal. For details about the APC please see here. Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

Chemoinformatics tools
Bioinformatics tools
Structure–activity modeling
Structure–property modeling
QSAR, QSPR
Drug/structure identification in bioprospecting research
Development of algorithms
Databases and web servers
Data mining
Structure representation and optimization
Molecular descriptors
Modelling of health and environmentally relevant endpoints/activities/properties
Toxicity, carcinogenicity
Computational methods
Model validation approaches
Multivariate modeling
Predictive modeling
Descriptive modeling
Classification modeling
Machine learning
Deep learning
Structure visualization

Published Papers (11 papers)

Download All Papers

Research

Jump to: Review

11 pages, 278 KiB

Open AccessArticle

Unsupervised Representation Learning for Proteochemometric Modeling

by Paul T. Kim, Robin Winter and Djork-Arné Clevert

Int. J. Mol. Sci. 2021, 22(23), 12882; https://0-doi-org.brum.beds.ac.uk/10.3390/ijms222312882 - 28 Nov 2021

Cited by 4 | Viewed by 3325

Abstract

In silico protein–ligand binding prediction is an ongoing area of research in computational chemistry and machine learning based drug discovery, as an accurate predictive model could greatly reduce the time and resources necessary for the detection and prioritization of possible drug candidates. Proteochemometric modeling (PCM) attempts to create an accurate model of the protein–ligand interaction space by combining explicit protein and ligand descriptors. This requires the creation of information-rich, uniform and computer interpretable representations of proteins and ligands. Previous studies in PCM modeling rely on pre-defined, handcrafted feature extraction methods, and many methods use protein descriptors that require alignment or are otherwise specific to a particular group of related proteins. However, recent advances in representation learning have shown that unsupervised machine learning can be used to generate embeddings that outperform complex, human-engineered representations. Several different embedding methods for proteins and molecules have been developed based on various language-modeling methods. Here, we demonstrate the utility of these unsupervised representations and compare three protein embeddings and two compound embeddings in a fair manner. We evaluate performance on various splits of a benchmark dataset, as well as on an internal dataset of protein–ligand binding activities and find that unsupervised-learned representations significantly outperform handcrafted representations. Full article

(This article belongs to the Special Issue Chemoinformatics and Bioinformatics Tools in Structure-Activity Modelling in Molecular Sciences)

► Show Figures

Figure 1

14 pages, 4912 KiB

Open AccessArticle

Generation of Non-Nucleotide CD73 Inhibitors Using a Molecular Docking and 3D-QSAR Approach

by Swapnil P. Bhujbal and Jung-Mi Hah

Int. J. Mol. Sci. 2021, 22(23), 12745; https://0-doi-org.brum.beds.ac.uk/10.3390/ijms222312745 - 25 Nov 2021

Cited by 4 | Viewed by 2021

Abstract

Radiotherapy and chemotherapy are conventional cancer treatments. Around 60% of all patients who are diagnosed with cancer receive radio- or chemotherapy in combination with surgery during their disease. Only a few patients respond to the blockage of immune checkpoints alone, or in combination therapy, because their tumours might not be immunogenic. Under these circumstances, an increasing level of extracellular adenosine via the activation of ecto-5’-nucleotidase (CD73) and consequent adenosine receptor signalling is a typical mechanism that tumours use to evade immune surveillance. CD73 is responsible for the conversion of adenosine monophosphate to adenosine. CD73 is overexpressed in various tumour types. Hence, targetting CD73’s signalling is important for the reversal of adenosine-facilitated immune suppression. In this study, we selected a potent series of the non-nucleotide small molecule inhibitors of CD73. Molecular docking studies were performed in order to examine the binding mode of the inhibitors inside the active site of CD73 and 3D-QSAR was used to study the structure–activity relationship. The obtained CoMFA (q² = 0.844, ONC = 5, r² = 0.947) and CoMSIA (q² = 0.804, ONC = 4, r² = 0.954) models showed reasonable statistical values. The 3D-QSAR contour map analysis revealed useful structural characteristics that were needed to modify non-nucleotide small molecule inhibitors. We used the structural information from the overall docking and 3D-QSAR results to design new, potent CD73 non-nucleotide inhibitors. The newly designed CD73 inhibitors exhibited higher activity (predicted pIC₅₀) than the most active compound of all of the derivatives that were selected for this study. Further experimental studies are needed in order to validate the new CD73 inhibitors. Full article

(This article belongs to the Special Issue Chemoinformatics and Bioinformatics Tools in Structure-Activity Modelling in Molecular Sciences)

► Show Figures

Graphical abstract

23 pages, 12560 KiB

Open AccessArticle

Prediction Models for Agonists and Antagonists of Molecular Initiation Events for Toxicity Pathways Using an Improved Deep-Learning-Based Quantitative Structure–Activity Relationship System

by Yasunari Matsuzaka, Shin Totoki, Kentaro Handa, Tetsuyoshi Shiota, Kota Kurosaki and Yoshihiro Uesawa

Int. J. Mol. Sci. 2021, 22(19), 10821; https://0-doi-org.brum.beds.ac.uk/10.3390/ijms221910821 - 06 Oct 2021

Cited by 6 | Viewed by 2031

Abstract

In silico approaches have been studied intensively to assess the toxicological risk of various chemical compounds as alternatives to traditional in vivo animal tests. Among these approaches, quantitative structure–activity relationship (QSAR) analysis has the advantages that it is able to construct models to predict the biological properties of chemicals based on structural information. Previously, we reported a deep learning (DL) algorithm-based QSAR approach called DeepSnap-DL for high-performance prediction modeling of the agonist and antagonist activity of key molecules in molecular initiating events in toxicological pathways using optimized hyperparameters. In the present study, to achieve high throughput in the DeepSnap-DL system–which consists of the preparation of three-dimensional molecular structures of chemical compounds, the generation of snapshot images from the three-dimensional chemical structures, DL, and statistical calculations—we propose an improved DeepSnap-DL approach. Using this improved system, we constructed 59 prediction models for the agonist and antagonist activity of key molecules in the Tox21 10K library. The results indicate that modeling of the agonist and antagonist activity with high prediction performance and high throughput can be achieved by optimizing suitable parameters in the improved DeepSnap-DL system. Full article

(This article belongs to the Special Issue Chemoinformatics and Bioinformatics Tools in Structure-Activity Modelling in Molecular Sciences)

► Show Figures

Figure 1

25 pages, 10258 KiB

Open AccessArticle

Molecular Dynamics-Derived Pharmacophore Model Explaining the Nonselective Aspect of K_V10.1 Pore Blockers

by Žan Toplak, Franci Merzel, Luis A. Pardo, Lucija Peterlin Mašič and Tihomir Tomašič

Int. J. Mol. Sci. 2021, 22(16), 8999; https://0-doi-org.brum.beds.ac.uk/10.3390/ijms22168999 - 20 Aug 2021

Cited by 3 | Viewed by 2031

Abstract

The K_V10.1 voltage-gated potassium channel is highly expressed in 70% of tumors, and thus represents a promising target for anticancer drug discovery. However, only a few ligands are known to inhibit K_V10.1, and almost all also inhibit the very similar cardiac hERG channel, which can lead to undesirable side-effects. In the absence of the structure of the K_V10.1–inhibitor complex, there remains the need for new strategies to identify selective K_V10.1 inhibitors and to understand the binding modes of the known K_V10.1 inhibitors. To investigate these binding modes in the central cavity of K_V10.1, a unique approach was used that allows derivation and analysis of ligand–protein interactions from molecular dynamics trajectories through pharmacophore modeling. The final molecular dynamics-derived structure-based pharmacophore model for the simulated K_V10.1–ligand complexes describes the necessary pharmacophore features for K_V10.1 inhibition and is highly similar to the previously reported ligand-based hERG pharmacophore model used to explain the nonselectivity of K_V10.1 pore blockers. Moreover, analysis of the molecular dynamics trajectories revealed disruption of the π–π network of aromatic residues F359, Y464, and F468 of K_V10.1, which has been reported to be important for binding of various ligands for both K_V10.1 and hERG channels. These data indicate that targeting the K_V10.1 channel pore is also likely to result in undesired hERG inhibition, and other potential binding sites should be explored to develop true K_V10.1-selective inhibitors as new anticancer agents. Full article

(This article belongs to the Special Issue Chemoinformatics and Bioinformatics Tools in Structure-Activity Modelling in Molecular Sciences)

► Show Figures

Figure 1

17 pages, 15092 KiB

Open AccessArticle

Prediction of Drug-Induced Liver Toxicity Using SVM and Optimal Descriptor Sets

by Keerthana Jaganathan, Hilal Tayara and Kil To Chong

Int. J. Mol. Sci. 2021, 22(15), 8073; https://0-doi-org.brum.beds.ac.uk/10.3390/ijms22158073 - 28 Jul 2021

Cited by 17 | Viewed by 3648

Abstract

Drug-induced liver toxicity is one of the significant safety challenges for the patient’s health and the pharmaceutical industry. It causes termination of drug candidates in clinical trials and also the retractions of approved drugs from the market. Thus, it is essential to identify hepatotoxic compounds in the initial stages of drug development process. The purpose of this study is to construct quantitative structure activity relationship models using machine learning algorithms and systematical feature selection methods for molecular descriptor sets. The models were built from a large and diverse set of 1253 drug compounds and were validated internally with 10-fold cross-validation. In this study, we applied a variety of feature selection techniques to extract the optimal subset of descriptors as modeling features to improve the prediction performance. Experimental results suggested that the support vector machine-based classifier had achieved a better classification accuracy with reduced molecular descriptors. The final optimal model provides an accuracy of 0.811, a sensitivity of 0.840, a specificity of 0.783 and Mathew’s correlation coefficient of 0.623 with an internal validation set. Furthermore, this model outperformed the prior studies while evaluated in both the internal and external test sets. The utilization of distinct optimal molecular descriptors as modeling features produce an in silico model with a superior performance. Full article

(This article belongs to the Special Issue Chemoinformatics and Bioinformatics Tools in Structure-Activity Modelling in Molecular Sciences)

► Show Figures

Figure 1

26 pages, 5186 KiB

Open AccessArticle

Biological Activities Related to Plant Protection and Environmental Effects of Coumarin Derivatives: QSAR and Molecular Docking Studies

by Vesna Rastija, Karolina Vrandečić, Jasenka Ćosić, Ivana Majić, Gabriella Kanižai Šarić, Dejan Agić, Maja Karnaš, Melita Lončarić and Maja Molnar

Int. J. Mol. Sci. 2021, 22(14), 7283; https://0-doi-org.brum.beds.ac.uk/10.3390/ijms22147283 - 06 Jul 2021

Cited by 10 | Viewed by 2980

Abstract

The aim was to study the inhibitory effects of coumarin derivatives on the plant pathogenic fungi, as well as beneficial bacteria and nematodes. The antifungal assay was performed on four cultures of phytopathogenic fungi by measuring the radial growth of the fungal colonies. Antibacterial activity was determined by the broth microdilution method performed on two beneficial soil organisms. Nematicidal activity was tested on two entomopathogenic nematodes. The quantitative structure-activity relationship (QSAR) model was generated by genetic algorithm, and toxicity was estimated by T.E.S.T. software. The mode of inhibition of enzymes related to the antifungal activity is elucidated by molecular docking. Coumarin derivatives were most effective against Macrophomina phaseolina and Sclerotinia sclerotiorum, but were not harmful against beneficial nematodes and bacteria. A predictive QSAR model was obtained for the activity against M. phaseolina (R²_tr = 0.78; R²_ext = 0.67; Q²_loo = 0.67). A QSAR study showed that multiple electron-withdrawal groups, especially at position C-3, enhanced activities against M. phaseolina, while the hydrophobic benzoyl group at the pyrone ring, and –Br, –OH, –OCH₃, at the benzene ring, may increase inhibition of S. sclerotiourum. Tested compounds possibly act inhibitory against plant wall-degrading enzymes, proteinase K. Coumarin derivatives are the potentially active ingredient of environmentally friendly plant-protection products. Full article

(This article belongs to the Special Issue Chemoinformatics and Bioinformatics Tools in Structure-Activity Modelling in Molecular Sciences)

► Show Figures

Figure 1

13 pages, 1385 KiB

Open AccessArticle

Combined Naïve Bayesian, Chemical Fingerprints and Molecular Docking Classifiers to Model and Predict Androgen Receptor Binding Data for Environmentally- and Health-Sensitive Substances

by Alfonso T. García-Sosa and Uko Maran

Int. J. Mol. Sci. 2021, 22(13), 6695; https://0-doi-org.brum.beds.ac.uk/10.3390/ijms22136695 - 22 Jun 2021

Cited by 3 | Viewed by 2105

Abstract

Many chemicals that enter the environment, food chain, and the human body can disrupt androgen-dependent pathways and mimic hormones and therefore, may be responsible for multiple diseases from reproductive to tumor. Thus, modeling and predicting androgen receptor activity is an important area of research. The aim of the current study was to find a method or combination of methods to predict compounds that can bind to and/or disrupt the androgen receptor, and thereby guide decision making and further analysis. A stepwise procedure proceeded from analysis of protein structures from human, chimp, and rat, followed by docking and subsequent ligand, and statistics based techniques that improved classification gradually. The best methods used multivariate logistic regression of combinations of chimpanzee protein structural docking scores, extended connectivity fingerprints, and naïve Bayesians of known binders and non-binders. Combination or consensus methods included data from a variety of procedures to improve the final model accuracy. Full article

(This article belongs to the Special Issue Chemoinformatics and Bioinformatics Tools in Structure-Activity Modelling in Molecular Sciences)

► Show Figures

Figure 1

13 pages, 3361 KiB

Open AccessArticle

Evaluation of Selective COX-2 Inhibition and In Silico Study of Kuwanon Derivatives Isolated from Morus alba

by Seung-Hwa Baek, Sungbo Hwang, Tamina Park, Yoon-Ju Kwon, Myounglae Cho and Daeui Park

Int. J. Mol. Sci. 2021, 22(7), 3659; https://0-doi-org.brum.beds.ac.uk/10.3390/ijms22073659 - 01 Apr 2021

Cited by 17 | Viewed by 2480

Abstract

Six kuwanon derivatives (A/B/C/E/H/J) extracted from the roots of Morus alba L. were evaluated to determine their cyclooxygenase (COX)-1 and 2 inhibitory effects. Cyclooxygenase (COX) is known as the target enzyme of nonsteroidal anti-inflammatory drugs (NSAIDs), which are the most widely used therapeutic agents for pain and inflammation. Among six kuwanon derivatives, kuwanon A showed selective COX-2 inhibitory activity, almost equivalent to that of celecoxib, a known COX inhibitor. Kuwanon A showed high COX-2 inhibitory activity (IC₅₀ = 14 μM) and a selectivity index (SI) range of >7.1, comparable to celecoxib (SI > 6.3). To understand the mechanisms underlying this effect, we performed docking simulations, fragment molecular orbital (FMO) calculations, and pair interaction energy decomposition analysis (PIEDA) at the quantum-mechanical level. As a result, kuwanon A had the strongest interaction with Arg120 and Tyr355 at the gate of the COX active site (−7.044 kcal/mol) and with Val89 in the membrane-binding domain (−6.599 kcal/mol). In addition, kuwanon A closely bound to Val89, His90, and Ser119, which are residues at the entrance and exit routes of the COX active site (4.329 Å). FMO calculations and PIEDA well supported the COX-2 selective inhibitory action of kuwanon A. It showed that the simulation and modeling results and experimental evidence were consistent. Full article

(This article belongs to the Special Issue Chemoinformatics and Bioinformatics Tools in Structure-Activity Modelling in Molecular Sciences)

► Show Figures

Figure 1

15 pages, 1903 KiB

Open AccessArticle

Structure-Activity Relationship Modeling and Experimental Validation of the Imidazolium and Pyridinium Based Ionic Liquids as Potential Antibacterials of MDR Acinetobacter baumannii and Staphylococcus aureus

by Ivan V. Semenyuta, Maria M. Trush, Vasyl V. Kovalishyn, Sergiy P. Rogalsky, Diana M. Hodyna, Pavel Karpov, Zhonghua Xia, Igor V. Tetko and Larisa O. Metelytsia

Int. J. Mol. Sci. 2021, 22(2), 563; https://0-doi-org.brum.beds.ac.uk/10.3390/ijms22020563 - 08 Jan 2021

Cited by 10 | Viewed by 2550

Abstract

Online Chemical Modeling Environment (OCHEM) was used for QSAR analysis of a set of ionic liquids (ILs) tested against multi-drug resistant (MDR) clinical isolate Acinetobacter baumannii and Staphylococcus aureus strains. The predictive accuracy of regression models has coefficient of determination q² = 0.66 − 0.79 with cross-validation and independent test sets. The models were used to screen a virtual chemical library of ILs, which was designed with targeted activity against MDR Acinetobacter baumannii and Staphylococcus aureus strains. Seven most promising ILs were selected, synthesized, and tested. Three ILs showed high activity against both these MDR clinical isolates. Full article

(This article belongs to the Special Issue Chemoinformatics and Bioinformatics Tools in Structure-Activity Modelling in Molecular Sciences)

► Show Figures

Figure 1

Review

Jump to: Research

21 pages, 2571 KiB

Open AccessReview

Moving Average-Based Multitasking In Silico Classification Modeling: Where Do We Stand and What Is Next?

by Amit Kumar Halder, Ana S. Moura and Maria Natália D. S. Cordeiro

Int. J. Mol. Sci. 2022, 23(9), 4937; https://0-doi-org.brum.beds.ac.uk/10.3390/ijms23094937 - 29 Apr 2022

Cited by 7 | Viewed by 1620

Abstract

Conventional in silico modeling is often viewed as ‘one-target’ or ‘single-task’ computer-aided modeling since it mainly relies on forecasting an endpoint of interest from similar input data. Multitasking or multitarget in silico modeling, in contrast, embraces a set of computational techniques that efficiently integrate multiple types of input data for setting up unique in silico models able to predict the outcome(s) relating to various experimental and/or theoretical conditions. The latter, specifically, based upon the Box–Jenkins moving average approach, has been applied in the last decade to several research fields including drug and materials design, environmental sciences, and nanotechnology. The present review discusses the current status of multitasking computer-aided modeling efforts, meanwhile describing both the existing challenges and future opportunities of its underlying techniques. Some important applications are also discussed to exemplify the ability of multitasking modeling in deriving holistic and reliable in silico classification-based models as well as in designing new chemical entities, either through fragment-based design or virtual screening. Focus will also be given to some software recently developed to automate and accelerate such types of modeling. Overall, this review may serve as a guideline for researchers to grasp the scope of multitasking computer-aided modeling as a promising in silico tool. Full article

(This article belongs to the Special Issue Chemoinformatics and Bioinformatics Tools in Structure-Activity Modelling in Molecular Sciences)

► Show Figures

Figure 1

20 pages, 2191 KiB

Open AccessReview

Towards Quantum-Chemical Modeling of the Activity of Anesthetic Compounds

by Janusz Cukras and Joanna Sadlej

Int. J. Mol. Sci. 2021, 22(17), 9272; https://0-doi-org.brum.beds.ac.uk/10.3390/ijms22179272 - 27 Aug 2021

Cited by 3 | Viewed by 2690

Abstract

The modeling of the activity of anesthetics is a real challenge because of their unique electronic and structural characteristics. Microscopic approaches relevant to the typical features of these systems have been developed based on the advancements in the theory of intermolecular interactions. By stressing the quantum chemical point of view, here, we review the advances in the field highlighting differences and similarities among the chemicals within this group. The binding of the anesthetics to their partners has been analyzed by Symmetry-Adapted Perturbation Theory to provide insight into the nature of the interaction and the modeling of the adducts/complexes allows us to rationalize their anesthetic properties. A new approach in the frame of microtubule concept and the importance of lipid rafts and channels in membranes is also discussed. Full article

(This article belongs to the Special Issue Chemoinformatics and Bioinformatics Tools in Structure-Activity Modelling in Molecular Sciences)

► Show Figures