Carlson, Heather A, Kevin M Masukawa, and J Andrew McCammon. “Method for Including the Dynamic Fluctuations of a Protein in Computer-Aided Drug Design.” The Journal of Physical Chemistry A 103, no. 49 (1999): 10213–19.

Two dynamic pharmacophore models were created. One used methanol probe molecules in a molecular dynamics simulation to find the active site and pharmacophoric features like hydrogen bond donors and acceptors. The other model was created from two crystal structures of the target protein. Both models got better results than pharmacophores derived from static receptors.

Wieder, Marcus, Ugo Perricone, Thomas Seidel, and Thierry Langer. “Pharmacophore Models Derived from Molecular Dynamics Simulations of Protein-Ligand Complexes: A Case Study.” Natural Product Communications 11, no. 10 (2016): 1934578X1601101019.

A study that applied molecular dynamics simulation to generate pharmacophores for insulin growth factor. The frequency of the features that appeared in the MD simulations was then used to generate two distinct pharmacophore models. It was shown that MD simulations can reveal otherwise hidden pharmacophore features that are not present in the pharmacophore model derived from the experimental crystal structure.

Duan, Jianxin, Steven L. Dixon, Jeffrey F. Lowrie, and Woody Sherman. “Analysis and Comparison of 2D Fingerprints: Insights into Database Screening Performance Using Eight Fingerprint Methods.” Journal of Molecular Graphics and Modelling 29, no. 2 (2010): 157–70.

This paper compared different methods to produce fingerprints. Each method was used for virtual screening and their performance was compared. It was concluded that there is not a better method and depending on the problem a particular method can be useful. Also, some metrics for evaluating virtual screening results using fingerprints are mentioned.

Desaphy, Jérémy, Karima Azdimousa, Esther Kellenberger, and Didier Rognan. “Comparison and Druggability Prediction of Protein–Ligand Binding Sites from Pharmacophore-Annotated Cavity Shapes.” Journal of Chemical Information and Modeling 52, no. 8 (2012): 2287–99.

An approach to describe the cavity of a protein using pharmacophoric points is proposed. This method can be used to compare the binding sites of different proteins and measure the similarity as well as predict the druggability of a protein binding site.

Koes, David Ryan, and Carlos J Camacho. “Pharmer: Efficient and Exact Pharmacophore Search.” Journal of Chemical Information and Modeling 51, no. 6 (2011): 1307–14.

The article presents a software named Pharmer that can be used to screen databases with pharmacophore queries. The novelty is that this software uses a type of data structures named KDB-tree and Bloom fingerprints which make the search much faster compared to other software.

Tunca, Guzin. “A Virtual Screening Procedure Combining Pharmacophore Filtering and Molecular Docking with the LIE Method,” 2012.

In the thesis the authors propose a virtual screening procedure that combines pharmacophores, docking, molecular dynamics, and calculation of binding free energy via a method called LIE. Pharmacophores can be used to filter large databases that then can be used for docking, after scoring only a few compounds are left for which the binding free energies are estimated to select the ones to be used for experimental tests.

The authors made emphasis that drug discovery is a such a complex problem and approaches combining different techniques are needed to compensate for the weaknesses of one another. Thus, pharmacophore modeling followed by docking and molecular dynamics can be a good methodology.

Wolber, Gerhard, and Robert Kosara. “Pharmacophores and Pharmacophore Searches.” Methods and Principles in Medicinal Chemistry, 2016, 131–50.

Explains in detail how pharmacophore features such as hydrogen bond donors/acceptors, hydrophobic areas, and positive and negative ionizable areas are defined by Ligand Scout. Moreover, it describes a classification of features based on specificity.

Additionally, a preprocessing method for PDB structures is shown. It consists in determining hybridization states, double bonds, rings, etc.

The paper describes an algorithm for finding rings in a molecular graph. It is based on the breadth first search algorithm.
This algorithm could be used for finding aromatic features in a pharmacophore model.

Wolber, Gerhard, and Thierry Langer. “LigandScout: 3-D Pharmacophores Derived from Protein-Bound Ligands and Their Use as Virtual Screening Filters.” Journal of Chemical Information and Modeling 45, no. 1 (2005): 160–69.

Explains the methods used by the Ligand Scout software. First, PDB files need to be treated to determine hybridization states, functional groups, and double bonds. Next, features such as hydrogen bond interactions are chosen based on a certain tolerance like the distance and angles. All features are represented by a sphere of 1.5 Å radius. Finally, different pharmacophore models are generated and then aligned to form one that can be used for virtual screening.

Paper also explains how pharmacophores were generated for Human Rhinovirus Serotype 16 Inhibitors and Tyrosine Kinase Inhibitors.

Shoichet, Brian K, Susan L McGovern, Binqing Wei, and John J Irwin. “Lead Discovery Using Molecular Docking.” Current Opinion in Chemical Biology 6, no. 4 (2002): 439–46.

The article compares docking with high-throughput screening (HTS), showing that in many cases virtual screening using docking has had better results than HTS. Moreover, it presents some ligands that were discovered through docking.

Finally, it mentions some issues with docking like dealing with receptor flexibility, errors in scoring functions and finding promiscuous or non-drug-like inhibitors.

Lupia, Antonio, Federica Moraca, Donatella Bagetta, Annalisa Maruca, Francesca Alessandra Ambrosio, Roberta Rocca, Raffaella Catalano, et al. “Computer-Based Techniques for Lead Identification and Optimization II: Advanced Search Methods.” Physical Sciences Reviews 5, no. 5 (2019): 20180114.

The paper discusses a technique called Metadynamics which can be used to estimate the binding free energy, it mentions the equations and procedure for this technique as well as a case study.

It also describes pharmacophores obtained from molecular dynamics simulations, otherwise known as dynophores. Finally, A procedure to find a representative pharmacophore model from a MD simulation is explained.

Sydow, Dominique. “Dynophores: Novel Dynamic Pharmacophores,” PhD Thesis, 2015.

Using molecular dynamics, a richer pharmacophore model that includes information about flexibility and conformational space of the molecules can be created. This is called a dynophore. The thesis discusses the application of dynophores and how they are created employing the DynophoreApp software.

The DynophporeApp was created using Java and the LiganScout framework, and the thesis goes into some detail of how it was implemented (it describes classes that were used, etc.).

Durairaj, Pradeepraj, Linbing Fan, David Machalz, Gerhard Wolber, and Matthias Bureik. “Functional Characterization and Mechanistic Modeling of the Human Cytochrome P450 Enzyme CYP4A22.” FEBS Letters 593, no. 16 (2019): 2214–25.

A study of an enzyme, a human cytochrome (CYP4A22), that was believed to be inactive. In the article, the authors demonstrated that the protein indeed has catalytic activity. With the use of dynamic pharmacophores, the residues important for ligand binding were discovered. Furthermore, docking was used to confirm which ligands bind to the enzyme.

Erickson, Jon A., Mehran Jalaie, Daniel H. Robertson, Richard A. Lewis, and Michal Vieth. “Lessons in Molecular Recognition: The Effects of Ligand and Protein Flexibility on Molecular Docking Accuracy.” Journal of Medicinal Chemistry 47, no. 1 (2004): 45–55.

A case study that evaluated different docking algorithms and their ability to successfully dock ligands. The docking algorithms used did not consider protein flexibility. However, good scores were obtained, especially with a program called CDOCKER which utilizes molecular dynamics for ligand flexibility.

The study explained that one of the major problems of docking is ligand and protein flexibility. In fact, when the number of rotatable bonds increased for the ligand, the algorithms success rate decreased. Similarly, it was shown that docking accuracy falls substantially when using an average structure and that it is correlated with the degree of protein movement.

Wieder, Marcus, Arthur Garon, Ugo Perricone, Stefan Boresch, Thomas Seidel, Anna Maria Almerico, and Thierry Langer. “Common Hits Approach: Combining Pharmacophore Modeling and Molecular Dynamics Simulations.” Journal of Chemical Information and Modeling 57, no. 2 (2017): 365–85.

An approach to improve the results from virtual screening using pharmacophores obtained from molecular dynamics simulations is proposed. This is called the “Common Hits Approach” and it consists of extracting representative pharmacophores from MD simulations, screening the database with each one and scoring retrieved molecules by counting how many times they are retrieved with each pharmacophore model.

The results in the study show that this approach is superior to using the pharmacophore model from the protein structure without MD, and it is also better than choosing the pharmacophore model that appears the most in the MD simulations.

Hu, Bingjie, and Markus A Lill. “PharmDock: A Pharmacophore-Based Docking Program.” Journal of Cheminformatics 6, no. 1 (2014): 14.

PharmDock is a program that generates structured based pharmacophores, performs docking and estimates the binding free energy. The article explains how the different phases of PharmDock were implemented. It mentions the algorithms and functions used to create pharmacophores, the sampling of the ligands and docking.

Part of PharmDock was built on python and it uses PyMol to visualize molecules.

Amaro, Rommie E., Jerome Baudry, John Chodera, Özlem Demir, J. Andrew McCammon, Yinglong Miao, and Jeremy C. Smith. “Ensemble Docking in Drug Discovery.” Biophysical Journal 114, no. 10 (2018): 2271–78.

One of the limitations of molecular dynamics is the sampling problem, which refers to the insufficient sampling of target configurational space, due to a large gap between the timescales reachable by simulation (usually microseconds) and slow target conformational changes, which can be many orders of magnitude longer. One method that could potentially solve this is Markov State models.

Khedkar, Santosh A., Alpeshkumar K. Malde, Evans C. Coutinho, and Sudha Srivastava. "Pharmacophore modeling in drug discovery and development: an overview." Medicinal Chemistry 3, no. 2 (2007): 187-197.

Overview of pharmacophore models. Describes the process of virtual screening from end to end.
Introduces the fingerprint concept as well as some applications of pharmacophores. Mentions some of the algorithms and software that are currently being used.

[by @Daniel-Ibarrola]

Wieder, Marcus, Ugo Perricone, Thomas Seidel, Stefan Boresch, and Thierry Langer. “Comparing Pharmacophore Models Derived from Crystal Structures and from Molecular Dynamics Simulations.” Monatshefte Für Chemie - Chemical Monthly 147, no. 3 (2016): 553–63.

Pharmacophore models obtained from the last frame of a MD simulation were compared with models obtained from the crystal structure. It was observed that the two models produce different results and that using information obtained from MD can improve pharmacophores.

Leach, Andrew R., Valerie J. Gillet, Richard A. Lewis, and Robin Taylor. “Three-Dimensional Pharmacophore Methods in Drug Discovery.” Journal of Medicinal Chemistry 53, no. 2 (2010): 539–58.

Discusses features, algorithms used, and scoring functions for pharmacophore generation. Also, it covers how databases are searched and created. Moreover, it includes a small discussion of field-based pharmacophores, and evaluation methods. Article discusses some of the main issues associated with pharmacophores including problems with molecular alignment, the need for new validation standards, difficulties when deciding which features to use, (e.g., which atoms should be considered hydrophobic), and that variable binding modes should be considered.

Nizami, Bilal, Dominique Sydow, Gerhard Wolber, and Bahareh Honarparvar. “Molecular Insight on the Binding of NNRTI to K103N Mutated HIV-1 RT: Molecular Dynamics Simulations and Dynamic Pharmacophore Analysis.” Molecular BioSystems 12, no. 11 (2016): 3385–95.

A comparison between the mutated reverse transcriptase of HIV and the wild type was carried out using molecular dynamics simulations. Different metrics such as RMSD, radius of gyration and a PCA analysis were computed to investigate the difference in conformation of the two proteins. Finally, dynophores were used to analyze the binding mode of an inhibitor on the two enzymes.

Koes, David Ryan, and Carlos J. Camacho. “ZINCPharmer: Pharmacophore Search of the ZINC Database.” Nucleic Acids Research 40, no. W1 (2012): W409–14.

ZincPharmer is a web application that uses the ZINC database for virtual screening. It is based on Pharmer so it can search the database faster. The article explains some of the features of ZincPharmer and how it works.

Walters, W.Patrick, Matthew T Stahl, and Mark A Murcko. “Virtual Screening—an Overview.” Drug Discovery Today 3, no. 4 (1998): 160–78.

A review of virtual screening process, including applications of pharmacophores and docking.

Despite being 20 years old, the article mentions the same problems associated with virtual screening and that can also be found in pharmacophores and docking:

• Protein Flexibility
• Water molecules
• Scoring functions
• Desolvation
• Entropy
• Binding Free energy calculations

Zoe Cournia, Bryce K. Allen, Thijs Beuming, David A. Pearlman, Brian K. Radak, and Woody Sherman. "Rigorous free energy simulations in virtual screening". Journal of Chemical Information and Modeling 60, no. 9 (2020): 4153-4169.

This article covers virtual screening methods including ligand-based methods and docking.
It also gives a review of the thermodynamics of binding and different methods to calculate binding free energy.

[by @Daniel-Ibarrola ]

Maruca, Annalisa, Francesca Alessandra Ambrosio, Antonio Lupia, Isabella Romeo, Roberta Rocca, Federica Moraca, Carmine Talarico, et al. “Computer-Based Techniques for Lead Identification and Optimization I: Basics.” Physical Sciences Reviews 4, no. 6 (2019): 20180113.

An overview of virtual screening, pharmacophores, docking, and molecular dynamics simulations. It mentions methods for validation of pharmacophores such as sensitivity, specificity, enrichment factor, etc

Schaller, David, Szymon Pach, Marcel Bermudez, and Gerhard Wolber. “Protein-Ligand Interactions and Drug Design.” Methods in Molecular Biology 2266 (2021): 227–38.

PyRod is a python-based library to generate three-dimensional pharmacophore models based on water traces of molecular dynamics simulations. The article describes a workflow using PyRod to generate pharmacophores for virtual screening. Nevertheless, the proposed protocol requires the use of different software including commercial one such as Ligand Scout

Yang, Sheng-Yong. "Pharmacophore modeling and applications in drug discovery: challenges and recent advances." Drug discovery today 15, no. 11-12 (2010): 444-450.

Explains the pharmacophore concept and gives a short review of ligand-based, structured-based,
and virtual screening models as well as pharmacophore-based de novo design.

[by @Daniel-Ibarrola]

Dror, Oranit, Dina Schneidman-Duhovny, Yuval Inbar, Ruth Nussinov, and Haim J Wolfson. “Novel Approach for Efficient Pharmacophore-Based Virtual Screening: Method and Applications.” Journal of Chemical Information and Modeling 49, no. 10 (2009): 2333–43.

An article that covers a procedure for generating ligand-based pharmacophores. First, ligands are aligned in pairs and are scored. Then, multiple ligands are aligned, and their scores are computed. Finally, pharmacophoric features are chosen through a procedure in which weights are assigned to each one based on the frequency of appearance, these are called weighted pharmacophores.

Before virtual screening, the best pharmacophores are chosen by clustering the ones with similar pattern of features. The pharmacophores mentioned in the article were generated by a free software named PharmaGist.

Craig, Ian R, Jonathan W Essex, and Katrin Spiegel. “Ensemble Docking into Multiple Crystallographically Derived Protein Structures: An Evaluation Based on the Statistical Analysis of Enrichments.” Journal of Chemical Information and Modeling 50, no. 4 (2010): 511–24.

A comparison between standard docking and ensemble docking was made. It was found that only in some cases ensemble docking was superior to standard docking. Moreover, induced docking was also tested, finding that it produced better results.

[Falcon, Wilfredo Evangelista, Sally R Ellingson, Jeremy C Smith, and Jerome Baudry. “Ensemble Docking in Drug Discovery: How Many Protein Configurations from Molecular Dynamics Simulations Are Needed To Reproduce Known Ligand Binding?” The Journal of Physical Chemistry B 123, no. 25 (2019): 5189–95.] (https://doi.org/10.1021/acs.jpcb.8b11491)

Different protein conformations for docking were obtained from MD simulations. The article compared docking performance obtained by using all conformations obtained from MD simulations with docking performance achieved by only using a subset of the conformers. Performance using all conformations outperformed the results obtained by using the clustered conformations. However, the computing power required was much greater, and using only the clustered conformations can still provide good results.

Güner, O. F., & Bowen, J. P. (2014). Setting the record straight: the origin of the pharmacophore concept. Journal of chemical information and modeling, 54(5), 1269-1283.

This article covers virtual screening methods including ligand-based methods and docking.
It also gives a review of the thermodynamics of binding and different methods to calculate binding free energy.

[by @Daniel-Ibarrola]

Sunseri, Jocelyn, and David Ryan Koes. “Pharmit: Interactive Exploration of Chemical Space.” Nucleic Acids Research 44, no. W1 (2016): W442–48.

Pharmit is a virtual screening web app that allows searching different compound libraries as well as using a custom one. In Pharmit, databases can be screening using not only pharmacophores, but it also provides a shape-based screening method. Pharmit uses Pharmer technology to search database so it’s faster than most other virtual screening programs.

Polishchuk, Pavel, Alina Kutlushina, Dayana Bashirova, Olena Mokshyna, and Timur Madzhidov. “Virtual Screening Using Pharmacophore Models Retrieved from Molecular Dynamic Simulations.” International Journal of Molecular Sciences 20, no. 23 (2019): 5834.

A study that used molecular dynamics simulations to built pharmacophores. To build the pharmacophores, a python library named PharMD was developed.

Representative pharmacophores were selected using hashes instead of the traditional clustering algorithms. Finally, the authors propose a new approach for ranking pharmacophore models called the “conformers coverage approach” that does not need knowledge from active or inactive compounds.

Michelarakis, Nicholas. “Towards Dynamic Pharmacophore Models through the Use of Coarse Grained Molecular Dynamic Simulations,” PhD Thesis, 2019.

A thesis that describes the application of dynamic pharmacophores to different kinds of proteins including GPCR, lipid binding proteins and other more common ones. The authors used a molecular dynamics method called coarse graining which does not use an atomic resolution; however, it requires significantly less computing power and thus, the simulations can be run for a longer time.

The thesis also contains a good revision of molecular dynamics, explaining the fundamentals and the different interactions that are considered in the simulations.

Gaurav, Anand, and Vertika Gautam. “Structure-Based Three-Dimensional Pharmacophores as an Alternative to Traditional Methodologies.” Journal of Receptor, Ligand and Channel Research Volume 7 (2014): 27–38.

According to the authors structured-based pharmacophores are superior to ligand-base ones. This is because ligand models ignore intricate details of the binding site shape as well as interactions sites. Furthermore, structured-based pharmacophores followed by docking can provide superior results for finding new leads.

The article also offers a review of some of the most popular structured-based pharmacophore software. It states that the main weaknesses of these programs are that they cannot be applied in situations where only the target structure is available and that they lack the option to export pharmacophores to other programs.

Du, Wei, David Machalz, Qi Yan, Erik J. Sorensen, Gerhard Wolber, and Matthias Bureik. “Importance of Asparagine-381 and Arginine-487 for Substrate Recognition in CYP4Z1.” Biochemical Pharmacology 174 (2020): 113850.

The paper describes the investigation of a cytochrome enzyme which had not been studied extensively. With the use of molecular dynamics, they found new residues that were important for interaction with the enzyme’s substrate. Furthermore, dynamic pharmacophores helped elucidate the interaction patterns of the protein with its ligands.

I am not sure if any student will understand that the *.md files are plain text that is going to be read as Markdown. Maybe not every student knows Markdown at this point.

Horvath, Dragos. “Chemoinformatics and Computational Chemical Biology.” Methods in Molecular Biology 672 (2010): 261–98.

A review of different pharmacophore techniques including ligand-based, structured-based, and fingerprints and some of the main issues with each one. Some of the issues discussed are conformational sampling, feature detection, and protein flexibility. Also, many successful applications of pharmacophores for finding new hits are discussed.

Kutlushina, Alina, Aigul Khakimova, Timur Madzhidov, and Pavel Polishchuk. “Ligand-Based Pharmacophore Modeling Using Novel 3D Pharmacophore Signatures.” Molecules 23, no. 12 (2018): 3094.

Pmapper, an open-source software built with python, was developed to aid with ligand-based pharmacophore generation. The novelty is that it uses a different algorithm that does not rely on alignment of compounds to find a common pharmacophore.

In the article, the algorithm used in the software is described, as well as the targets and datasets used to validate these methods.

Rogers, David, and Mathew Hahn. “Extended-Connectivity Fingerprints.” Journal of Chemical Information and Modeling 50, no. 5 (2010): 742–54.

This paper discusses different fingerprint methods and its applications, it focuses on extended connectivity fingerprints but it also mentions functional connectivity fingerprints, which are used in pharmacophores. It also explains in detail an algorithm to generate extended connectivity fingerprints.

Schaller, David, Dora Šribar, Theresa Noonan, Lihua Deng, Trung Ngoc Nguyen, Szymon Pach, David Machalz, Marcel Bermudez, and Gerhard Wolber. "Next generation 3D pharmacophore modeling." Wiley Interdisciplinary Reviews: Computational Molecular Science 10, no. 4 (2020): e1468.

Some application case studies of pharmacophores. Explains shortly different approaches to integrate molecular dynamics
into pharmacophore models. It also briefly covers machine learning methods that are being used for pharmacophore design.
Contains a table of available software for pharmacophore modeling.

[by @Daniel-Ibarrola]

Carlson, Heather A. "Protein flexibility and drug design: how to hit a moving target." Current opinion in chemical biology 6, no. 4 (2002): 447-452.

Neglecting protein flexibility causes errors in computational models. The article discusses the basic theory behind protein and
small molecule interactions as well as methods to include protein flexibility.

[by @Daniel-Ibarrola]

Amaro, Rommie, and Wilfred Li. “Emerging Methods for Ensemble-Based Virtual Screening.” Current Topics in Medicinal Chemistry 10, no. 1 (2010): 3–13.

Proteins are not rigid, they do not exist in a single native conformation, but in an ensemble of states with different energies. With the use of molecular dynamics approaches an ensemble of these conformational states can be generated to enhance the virtual screening process. This include dynamic pharmacophores where a pharmacophore model is created from MD trajectories, and the relaxed complex scheme that uses snapshots from MD simulations to search ligand libraries.

There are some challenges and difficulties associated with MD simulations such as the inclusion of water molecules interactions, the total time of the simulations, and the need for new methods to predict the binding free energy.

Vlachakis, Dimitrios, Paraskevas Fakourelis, Vasileios Megalooikonomou, Christos Makris, and Sophia Kossida. “DrugOn: A Fully Integrated Pharmacophore Modeling and Structure Optimization Toolkit.” PeerJ 3 (2015): e725.

DrugOn is a free software for pharmacophore modeling. It offers a complete pipeline that begins with the preprocessing of PDB files, then it optimizes the receptor structure and finally, it generates the pharmacophore. The article describes all the steps that are used in this software

Sastry, Madhavi, Jeffrey F. Lowrie, Steven L. Dixon, and Woody Sherman. “Large-Scale Systematic Analysis of 2D Fingerprint Methods and Parameters to Improve Virtual Screening Enrichments.” Journal of Chemical Information and Modeling 50, no. 5 (2010): 771–84.

A comparison of different fingerprints and parameters that can be used in the software CANVAS. It was found that fingerprints and atom-typing schemes that encode more information tend to perform better. Moreover, fingerprints that use larger bit spaces result in better enrichments.

A program based on a genetic algorithm (GA) was developed for pharmacophore mapping. The paper explains how the algorithm was constructed and the results obtained in different datasets of ligands.

Perricone, Ugo, Marcus Wieder, Thomas Seidel, Thierry Langer, and Alessandro Padova. “The Use of Dynamic Pharmacophore in Computer-Aided Hit Discovery: A Case Study,” 1824:317–33. Methods in Molecular Biology, 2018..

A case study that shows how a dynamic pharmacophore model was built using molecular dynamics. It explains step by step from the selection of structures from the PDB to the screening of the libraries and the ranking of the models. It describes the software and techniques used in each stage of the process.

For each step different software was used. The pharmacophore model was constructed using LigandScout.

Elokely, Khaled M, and Robert J Doerksen. “Docking Challenge: Protein Sampling and Molecular Docking Performance.” Journal of Chemical Information and Modeling 53, no. 8 (2013): 1934–45.

Paper covers a methodology for docking. It begins by retrieving protein structures from PDB, followed by protein preparation to remove or add hydrogen atoms and adjust tautomeric and protonation states. Ligands are prepared in a similar fashion. Before the actual docking, different protein sampling methods are used: rigid receptors, soft receptors, induced fit, and conformational ensembles.

After analyzing docking results for different sampling techniques, it was concluded that the best approach depends on the target. For example, if there are a lot of crystal structures available ensemble docking could be better.

Finally, it is important to carefully analyze water molecules as they may be artifact molecules caused by the crystallization technique.

Drie, John H. Van. “History of 3D Pharmacophore Searching: Commercial, Academic and Open-Source Tools.” Drug Discovery Today: Technologies 7, no. 4 (2011): e255–62.

The article presents the history of pharmacophores and the tools and software that have been developed. It mentions different software for pharmacophore generation including commercial and open source. Also, tools for chemistry and conformer generation are mentioned.

Kim, Kyun-Hwan, Nam Doo Kim, and Baik-Lin Seong. "Pharmacophore-based virtual screening: a review of recent applications." Expert opinion on drug discovery 5, no. 3 (2010): 205-222.

Different applications of pharmacophores including cancer inhibitors, antiviral inhibitors like HIV and influenza, metabolic syndrome inhibitors, and Alzheimer's inhibitors.

[by @Daniel-Ibarrola]