FlexNovo: structure-based searching in large fragment spaces

Synthetically Feasible De Novo Molecular Design of Leads Based on a Reinforcement Learning Model: AI-Assisted Discovery of an Anti-IBD Lead Targeting CXCR4

Artificial intelligence (AI) de novo molecular generation provides leads with novel structures for drug discovery. However, the target affinity and synthesizability of the generated molecules present critical challenges for the successful application of AI technology. Therefore, we developed an advanced reinforcement learning model to bridge the gap between the theory of de novo molecular generation and the practical aspects of drug discovery. This model utilizes chemical reaction templates and commercially available building blocks as a starting point and employs forward reaction prediction to generate molecules, while real-time docking and drug-likeness predictions are conducted to ensure synthesizability and drug-likeness. We applied this model to design active molecules targeting the inflammation-related receptor CXCR4 and successfully prepared them according to the AI-proposed synthetic routes. Several molecules exhibited potent anti-CXCR4 and anti-inflammatory activity in subsequent in vitro and in vivo assays. The top-performing compound XVI alleviated symptoms related to inflammatory bowel disease and showed reasonable pharmacokinetic properties.

Emerging structure-based computational methods to screen the exploding accessible chemical space

Structure-based virtual screening can be a valuable approach to computationally select hit candidates based on their predicted interaction with a protein of interest. The recent explosion in the size of chemical libraries increases the chances of hitting high-quality compounds during virtual screening exercises but also poses new challenges as the number of chemically accessible molecules grows faster than the computing power necessary to screen them. We review here two novel approaches rapidly gaining in popularity to address this problem: machine learning-accelerated and synthon-based library screening. We summarize the results from seminal proof-of-concept studies, highlight the latest developments, and discuss limitations and future directions.

SpaceGrow: efficient shape-based virtual screening of billion-sized combinatorial fragment spaces

The growing size of make-on-demand chemical libraries is posing new challenges to cheminformatics. These ultra-large chemical libraries became too large for exhaustive enumeration. Using a combinatorial approach instead, the resource requirement scales approximately with the number of synthons instead of the number of molecules. This gives access to billions or trillions of compounds as so-called chemical spaces with moderate hardware and in a reasonable time frame. While extremely performant ligand-based 2D methods exist in this context, 3D methods still largely rely on exhaustive enumeration and therefore fail to apply. Here, we present SpaceGrow: a novel shape-based 3D approach for ligand-based virtual screening of billions of compounds within hours on a single CPU. Compared to a conventional superposition tool, SpaceGrow shows comparable pose reproduction capacity based on RMSD and superior ranking performance while being orders of magnitude faster. Result assessment of two differently sized subsets of the eXplore space reveals a higher probability of finding superior results in larger spaces highlighting the potential of searching in ultra-large spaces. Furthermore, the application of SpaceGrow in a drug discovery workflow was investigated in four examples involving G protein-coupled receptors (GPCRs) with the aim to identify compounds with similar binding capabilities and molecular novelty.

A flexible data-free framework for structure-based de novo drug design with reinforcement learning

Contemporary structure-based molecular generative methods have demonstrated their potential to model the geometric and energetic complementarity between ligands and receptors, thereby facilitating the design of molecules with favorable binding affinity and target specificity. Despite the introduction of deep generative models for molecular generation, the atom-wise generation paradigm that partially contradicts chemical intuition limits the validity and synthetic accessibility of the generated molecules. Additionally, the dependence of deep learning models on large-scale structural data has hindered their adaptability across different targets. To overcome these challenges, we present a novel search-based framework, 3D-MCTS, for structure-based de novo drug design. Distinct from prevailing atom-centric methods, 3D-MCTS employs a fragment-based molecular editing strategy. The fragments decomposed from small-molecule drugs are recombined under predefined retrosynthetic rules, offering improved drug-likeness and synthesizability, overcoming the inherent limitations of atom-based approaches. Leveraging multi-threaded parallel simulations combined with a real-time energy constraint-based pruning strategy, 3D-MCTS achieves remarkable efficiency. At a fixed computational cost, it outperforms other state-of-the-art (SOTA) methods by producing molecules with enhanced binding affinity. Furthermore, its fragment-based approach ensures the generation of more dependable binding conformations, exhibiting a success rate 43.6% higher than that of other SOTAs. This advantage becomes even more pronounced when handling targets that significantly deviate from the training dataset. 3D-MCTS is capable of achieving thirty times more hits with high binding affinity than traditional virtual screening methods, which demonstrates the superior ability of 3D-MCTS to explore chemical space. Moreover, the flexibility of our framework makes it easy to incorporate domain knowledge during the process, thereby enabling the generation of molecules with desirable pharmacophores and enhanced binding affinity. The adaptability of 3D-MCTS is further showcased in metalloprotein applications, highlighting its potential across various drug design scenarios.

Navigating large chemical spaces in early-phase drug discovery

The size of actionable chemical spaces is surging, owing to a variety of novel techniques, both computational and experimental. As a consequence, novel molecular matter is now at our fingertips that cannot and should not be neglected in early-phase drug discovery. Huge, combinatorial, make-on-demand chemical spaces with high probability of synthetic success rise exponentially in content, generative machine learning models go hand in hand with synthesis prediction, and DNA-encoded libraries offer new ways of hit structure discovery. These technologies enable to search for new chemical matter in a much broader and deeper manner with less effort and fewer financial resources. These transformational developments require new cheminformatics approaches to make huge chemical spaces searchable and analyzable with low resources, and with as little energy consumption as possible. Substantial progress has been made in the past years with respect to computation as well as organic synthesis. First examples of bioactive compounds resulting from the successful use of these novel technologies demonstrate their power to contribute to tomorrow's drug discovery programs. This article gives a compact overview of the state-of-the-art.

Galileo: Three-dimensional searching in large combinatorial fragment spaces on the example of pharmacophores

Fragment spaces are an efficient way to model large chemical spaces using a handful of small fragments and a few connection rules. The development of Enamine's REAL Space has shown that large spaces of readily available compounds may be created this way. These are several orders of magnitude larger than previous libraries. So far, searching and navigating these spaces is mostly limited to topological approaches. A way to overcome this limitation is optimization via metaheuristics which can be combined with arbitrary scoring functions. Here we present Galileo, a novel Genetic Algorithm to sample fragment spaces. We showcase Galileo in combination with a novel pharmacophore mapping approach, called Phariety, enabling 3D searches in fragment spaces. We estimate the effectiveness of the approach with a small fragment space. Furthermore, we apply Galileo to two pharmacophore searches in the REAL Space, detecting hundreds of compounds fulfilling a HSP90 and a FXIa pharmacophore.

Chemical space docking enables large-scale structure-based virtual screening to discover ROCK1 kinase inhibitors

With the ever-increasing number of synthesis-on-demand compounds for drug lead discovery, there is a great need for efficient search technologies. We present the successful application of a virtual screening method that combines two advances: (1) it avoids full library enumeration (2) products are evaluated by molecular docking, leveraging protein structural information. Crucially, these advances enable a structure-based technique that can efficiently explore libraries with billions of molecules and beyond. We apply this method to identify inhibitors of ROCK1 from almost one billion commercially available compounds. Out of 69 purchased compounds, 27 (39%) have K_i values < 10 µM. X-ray structures of two leads confirm their docked poses. This approach to docking scales roughly with the number of reagents that span a chemical space and is therefore multiple orders of magnitude faster than traditional docking.

New peptidomimetic rhodesain inhibitors with improved selectivity towards human cathepsins

Parasitic cysteine proteases such as rhodesain (TbCatL) from Trypanosoma brucei rhodesiense are relevant targets for developing new potential drugs against parasitic diseases (e. g. Human African Trypanosomiasis). Designing selective inhibitors for parasitic cathepsins can be challenging as they share high structural similarities with human cathepsins. In this paper, we describe the development of novel peptidomimetic rhodesain inhibitors by applying a structure-based de novo design approach and molecular docking protocols. The inhibitors with a new scaffold in P2 and P3 position display high selectivity towards trypanosomal rhodesain over human cathepsins L and B and high antitrypanosomal activity. Vinylsulfonate 2a has emerged as a potent rhodesain inhibitor (k_2nd = 883 • 10³ M^-1 s^-1) with single-digit nanomolar binding affinity (K_i = 9 nM) and more than 150-fold selectivity towards human cathepsins and it thus constitutes an interesting starting compound for the further development of selective drugs against Human African Trypanosomiasis.

Exploration of Ultralarge Compound Collections for Drug Discovery

Designing new medicines more cheaply and quickly is tightly linked to the quest of exploring chemical space more widely and efficiently. Chemical space is monumentally large, but recent advances in computer software and hardware have enabled researchers to navigate virtual chemical spaces containing billions of chemical structures. This review specifically concerns collections of many millions or even billions of enumerated chemical structures as well as even larger chemical spaces that are not fully enumerated. We present examples of chemical libraries and spaces and the means used to construct them, and we discuss new technologies for searching huge libraries and for searching combinatorially in chemical space. We also cover space navigation techniques and consider new approaches to de novo drug design and the impact of the "autonomous laboratory" on synthesis of designed compounds. Finally, we summarize some other challenges and opportunities for the future.

Theoretical Studies on the Molecular Properties, Toxicity, and Biological Efficacy of 21 New Chemical Entities

New chemical entities (NCEs) such as small molecules and antibody-drug conjugates have strong binding affinity for biological targets, which provide deep insights into structure-specific interactions for the design of future drugs. As structures of drugs increase in complexity, the importance of computational predictions comes into sharp focus. Knowledge of various computational tools enables us to predict the molecular properties, toxicity, and biological efficacy of the drugs and help the medicinal chemists to discover new drugs more efficiently. Newly approved drugs have higher affinities for proteins and nucleic acids and are applied for the treatment of human diseases. We have carried out the computational studies of 21 such NCEs, specifically small molecules and antibody-drug conjugates, and studied the biological efficacy of these drugs. Their bioactivity score and molecular and pharmacokinetic properties were evaluated using online computer software programs, viz., Molinspiration and Osiris Property Explorer. The SwissTargetPrediction tool was used for the efficient prediction of protein targets for the NCEs. The results indicated higher stability for the drug complexes due to a larger HOMO-LUMO gap. A high electrophilicity index reflects good electrophilic behavior and high reactivity of the drugs. Lipinski's ''rule of five'' indicated that most of the drug complexes are likely to be orally active. These drugs also showed non-mutagenic, non-tumorigenic, non-irritant, and non-effective reproductive behavior. We hope that these studies will provide an insight into molecular recognition and definitely help the medicinal chemists to design new drugs in future.

De novo molecular drug design benchmarking

De novo molecular design for drug discovery is a growing field. Deep neural networks (DNNs) are becoming more widespread in their use for machine learning models. As more DNN models are proposed for molecular design, benchmarking methods are crucial for the comparision and validation of these models. This review looks at recently proposed benchmarking methods Fréchet ChemNet Distance, GuacaMol and Molecular Sets (MOSES), and provides a commentary on their future potential applications in de novo molecular drug design and possible next steps for further validation of these benchmarking methods.

Shape-Based Descriptors for Efficient Structure-Based Fragment Growing

Structure-based fragment growing is one of the key techniques in fragment-based drug design. Fragment growing is commonly practiced based on structural and biophysical data. Computational workflows are employed to predict which fragment elaborations could lead to high-affinity binders. Several such workflows exist but many are designed to be long running noninteractive systems. Shape-based descriptors have been proven to be fast and perform well at virtual-screening tasks. They could, therefore, be applied to the fragment-growing problem to enable an interactive fragment-growing workflow. In this work, we describe and analyze the use of specific shape-based directional descriptors for the task of fragment growing. The performance of these descriptors that we call ray volume matrices (RVMs) is evaluated on two data sets containing protein-ligand complexes. While the first set focuses on self-growing, the second measures practical performance in a cross-growing scenario. The runtime of screenings using RVMs as well as their robustness to three dimensional perturbations is also investigated. Overall, it can be shown that RVMs are useful to prefilter fragment candidates. For up to 84% of the 3299 generated self-growing cases and for up to 66% of the 326 generated cross-growing cases, RVMs could create poses with less than 2 Å root-mean-square deviation to the crystal structure with average query speeds of around 30,000 conformations per second. This opens the door for fast explorative screenings of fragment libraries.

The next level in chemical space navigation: going far beyond enumerable compound libraries

Recent innovations have brought pharmacophore-driven methods for navigating virtual chemical spaces, the size of which can reach into the billions of molecules, to the fingertips of every chemist. There has been a paradigm shift in the underlying computational chemistry that drives chemical space search applications, incorporating intelligent reaction knowledge into their core so that they can readily deliver commercially available molecules as nearest neighbor hits from within giant virtual spaces. These vast resources enable medicinal chemists to execute rapid scaffold-hopping experiments, rapid hit expansion, and structure-activity relationship (SAR) exploitation in largely intellectual property (IP)-free territory and at unparalleled low cost.

Chemistry-driven Hit-to-lead Optimization Guided by Structure-based Approaches

For several decades, hit identification for drug discovery has been facilitated by developments in both fragment-based and high-throughput screening technologies. However, a major bottleneck in drug discovery projects continues to be the optimization of primary hits from screening campaigns in order to derive lead compounds. Computational chemistry or molecular modeling can play an important role during this hit-to-lead (H2L) stage by both suggesting putative optimizations and decreasing the number of compounds to be experimentally synthesized and evaluated. However, it is also crucial to consider the feasibility of organically synthesizing these virtually designed compounds. Furthermore, the generated molecules should have reasonable physicochemical properties and be medicinally relevant. This review focuses on chemistry-driven and structure-based computational methods that can be used to tackle the difficult problem of H2L optimization, with emphasis being placed on the strategy developed in our laboratory.

Recent advances on CDK inhibitors: An insight by means of in silico methods

The cyclin dependent kinases (CDKs) are a small family of serine/threonine protein kinases that can act as a potential therapeutic target in several proliferative diseases, including cancer. This short review is a survey on the more recent research progresses in the field achieved by using in silico methods. All the "armamentarium" available to the medicinal chemists (docking protocols and molecular dynamics, fragment-based, de novo design, virtual screening, and QSAR) has been employed to the discovery of new, potent, and selective inhibitors of cyclin dependent kinases. The results cited herein can be useful to understand the nature of the inhibitor-target interactions, and furnish an insight on the structural/molecular requirements necessary to achieve the required selectivity against cyclin dependent kinases over other types of kinases.

Evaluation of a semi-automated workflow for fragment growing

Intelligent Automatic Design (IADE) is an expert system developed at Novartis to identify nonclassical bioisosteres. In addition to bioisostere searching, one could also use IADE to grow a fragment bound to a protein. Here we report an evaluation of IADE as a tool for fragment growing. Three examples from the literature served as test cases. In all three cases, IADE generated close analogues of the published compounds and reproduced their crystallographic binding modes. This exercise validated the use of the IADE system for fragment growing. We have also gained experience in optimizing the performance of IADE for this type of application.

Disruption of intrinsic motions as a mechanism for enzyme inhibition

Clostridium difficile (C. diff) is one of the most common and most severe hospital-acquired infections; its consequences range from lengthened hospital stay to outright lethality. C. diff causes cellular damage through the action of two large toxins TcdA and TcdB. Recently, there has been increased effort toward developing antitoxin therapies, rather than antibacterial treatments, in hopes of mitigating the acquisition of drug resistance. To date, no analysis of the recognition mechanism of TcdA or TcdB has been attempted. Here, we use small molecule flexible docking followed by unbiased molecular dynamics to obtain a more detailed perspective on how inhibitory peptides, exemplified by two species HQSPWHH and EGWHAHT function. Using principal component analysis and generalized masked Delaunay analysis, an examination of the conformational space of TcdB in its apo form as well as forms bound to the peptides and UDP-Glucose was performed. Although both species inhibit by binding in the active site, they do so in two very different ways. The simulations show that the conformational space occupied by TcdB bound to the two peptides are quite different and provide valuable insight for the future design of toxin inhibitors and other enzymes that interact with their substrates through conformational capture mechanisms and thus work by the disruption of the protein's intrinsic motions.

Fragment-based design, docking, synthesis, biological evaluation and structure-activity relationships of 2-benzo/benzisothiazolimino-5-aryliden-4-thiazolidinones as cycloxygenase/lipoxygenase inhibitors

Balanced modulation of several targets is one of the current strategies for the treatment of multi-factorial diseases. Based on the knowledge of inflammation mechanisms, it was inferred that the balanced inhibition of cyclooxygenase-1/cyclooxygenase-2/lipoxygenase might be a promising approach for treatment of such a multifactorial disease state as inflammation. Detection of fragments responsible for interaction with enzyme's binding site provides the basis for designing new molecules with increased affinity and selectivity. A new chemoinformatics approach was proposed and applied to create a fragment library that was used to design novel inhibitors of cycloxygenase-1/cycloxygenase-2/lipoxygenase enzymes. Potential binding sites were elucidated by docking. Synthesis of novel compounds, and the in vitro/in vivo biological testing confirmed the results of computational studies. The benzothiazolyl moiety was proved to be of great significance for developing more potent inhibitors.

De novo drug design

Computer-assisted molecular design supports drug discovery by suggesting novel chemotypes and compound modifications for lead structure optimization. While the aspect of synthetic feasibility of the automatically designed compounds has been neglected for a long time, we are currently witnessing an increased interest in this topic. Here, we review state-of-the-art software for de novo drug design with a special emphasis on fragment-based techniques that generate druglike, synthetically accessible compounds. The importance of scoring functions that can be used to predict compound reactivity and potency is highlighted, and several promising solutions are discussed. Recent practical validation studies are presented that have already demonstrated that rule-based fragment assembly can result in novel synthesizable compounds with druglike properties and a desired biological activity.

De novo design: balancing novelty and confined chemical space

IMPORTANCE OF THE FIELD

De novo drug design serves as a tool for the discovery of new ligands for macromolecular targets as well as optimization of known ligands. Recently developed tools aim to address the multi-objective nature of drug design in an unprecedented manner.

AREAS COVERED IN THIS REVIEW

This article discusses recent advances in de novo drug design programs and accessory programs used to evaluate compounds post-generation.

WHAT THE READER WILL GAIN

The reader is introduced to the challenges inherent in de novo drug design and will become familiar with current trends in de novo design. Furthermore, the reader will be better prepared to assess the value of a tool, and be equipped to design more elegant tools in the future.

TAKE HOME MESSAGE

De novo drug design can assist in the efficient discovery of new compounds with a high affinity for a given target. The inclusion of existing chemoinformatic methods with current structure-based de novo design tools provides a means of enhancing the therapeutic value of these generated compounds.

Structure-based discovery of A2A adenosine receptor ligands

The recent determination of X-ray structures of pharmacologically relevant GPCRs has made these targets accessible to structure-based ligand discovery. Here we explore whether novel chemotypes may be discovered for the A_2A adenosine receptor, based on complementarity to its recently determined structure. The A_2A adenosine receptor signals in the periphery and the CNS, with agonists explored as anti-inflammatory drugs and antagonists explored for neurodegenerative diseases. We used molecular docking to screen a 1.4 million compound database against the X-ray structure computationally and tested 20 high-ranking, previously unknown molecules experimentally. Of these 35% showed substantial activity with affinities between 200 nM and 9 μM. For the most potent of these new inhibitors, over 50-fold specificity was observed for the A_2A versus the related A₁ and A₃ subtypes. These high hit rates and affinities at least partly reflect the bias of commercial libraries toward GPCR-like chemotypes, an issue that we attempt to investigate quantitatively. Despite this bias, many of the most potent new ligands were novel, dissimilar from known ligands, providing new lead structures for modulation of this medically important target.

LoFT: similarity-driven multiobjective focused library design

We present LoFT, a tool for focused combinatorial library design. LoFT provides a set of algorithms, constructing a focused library from a chemical fragment space under optimization of multiple design criteria. A weighted multiobjective scoring function based on physicochemical descriptors is employed for traversing the chemical search space. The new aspect of LoFT is that a similarity-driven product-based library design approach is provided on fragment level. For this reason the feature tree descriptor is incorporated for similarity comparison of library compounds to given bioactive molecules as well as for diversifying the resulting libraries. The feature tree descriptor abstracts the molecular graph to a tree structure where the nodes are labeled with physicochemical properties. For comparison, the nodes of two trees are mapped onto each other. This strictly hierarchical mechanism is suitable for the efficient comparison of chemical fragments, allowing the evaluation of the resulting products on fragment level without explicitly enumerating them. LoFT was validated, applying three different data sets. Starting with a random reagent selection, we optimized the libraries using maximum similarity to known bioactive molecules and iteratively adding further criteria. Moreover, we compared these results with data we obtained with FTrees-FS.

SHOP: receptor-based scaffold HOPping by GRID-based similarity searches

A new field-derived 3D method for receptor-based scaffold hopping, implemented in the software SHOP, is presented. Information from a protein-ligand complex is utilized to substitute a fragment of the ligand with another fragment from a database of synthetically accessible scaffolds. A GRID-based interaction profile of the receptor and geometrical descriptions of a ligand scaffold are used to obtain new scaffolds with different structural features and are able to replace the original scaffold in the protein-ligand complex. An enrichment study was successfully performed verifying the ability of SHOP to find known active CDK2 scaffolds in a database. Additionally, SHOP was used for suggesting new inhibitors of p38 MAP kinase. Four p38 complexes were used to perform six scaffold searches. Several new scaffolds were suggested, and the resulting compounds were successfully docked into the query proteins.

Novel inhibitors of anthrax edema factor

Several pathogenic bacteria produce adenylyl cyclase toxins, such as the edema factor (EF) of Bacillus anthracis. These disturb cellular metabolism by catalyzing production of excessive amounts of the regulatory molecule cAMP. Here, a structure-based method, where a 3D-pharmacophore that fit the active site of EF was constructed from fragments, was used to identify non-nucleotide inhibitors of EF. A library of small molecule fragments was docked to the EF-active site in existing crystal structures, and those with the highest HINT scores were assembled into a 3D-pharmacophore. About 10,000 compounds, from over 2.7 million compounds in the ZINC database, had a similar molecular framework. These were ranked according to their docking scores, using methodology that was shown to achieve maximum accuracy (i.e., how well the docked position matched the experimentally determined site for ATP analogues in crystal structures of the complex). Finally, 19 diverse compounds with the best AutoDock binding/docking scores were assayed in a cell-based assay for their ability to reduce cAMP secretion induced by EF. Four of the test compounds, from different structural groups, inhibited in the low micromolar range. One of these has a core structure common to phosphatase inhibitors previously identified by high-throughput assays of a diversity library. Thus, the fragment-based pharmacophore identified a small number of diverse compounds for assay, and greatly enhanced the selection process of advanced lead compounds for combinatorial design.

Concept of combinatorial de novo design of drug-like molecules by particle swarm optimization

We present a fast stochastic optimization algorithm for fragment-based molecular de novo design (COLIBREE, Combinatorial Library Breeding). The search strategy is based on a discrete version of particle swarm optimization. Molecules are represented by a scaffold, which remains constant during optimization, and variable linkers and side chains. Different linkers represent virtual chemical reactions. Side-chain building blocks were obtained from pseudo-retrosynthetic dissection of large compound databases. Here, ligand-based design was performed using chemically advanced template search (CATS) topological pharmacophore similarity to reference ligands as fitness function. A weighting scheme was included for particle swarm optimization-based molecular design, which permits the use of many reference ligands and allows for positive and negative design to be performed simultaneously. In a case study, the approach was applied to the de novo design of potential peroxisome proliferator-activated receptor subtype-selective agonists. The results demonstrate the ability of the technique to cope with large combinatorial chemistry spaces and its applicability to focused library design. The technique was able to perform exploitation of a known scheme and at the same time explorative search for novel ligands within the framework of a given molecular core structure. It thereby represents a practical solution for compound screening in the early hit and lead finding phase of a drug discovery project.

Exploring fragment spaces under multiple physicochemical constraints

We present a new algorithm for the enumeration of chemical fragment spaces under constraints. Fragment spaces consist of a set of molecular fragments and a set of rules that specifies how fragments can be combined. Although fragment spaces typically cover an infinite number of molecules, they can be enumerated in case that a physicochemical profile of the requested compounds is given. By using min-max ranges for a number of corresponding properties, our algorithm is able to enumerate all molecules which obey these properties. To speed up the calculation, the given ranges are used directly during the build-up process to guide the selection of fragments. Furthermore, a topology based fragment filter is used to skip most of the redundant fragment combinations. We applied the algorithm to 40 different target classes. For each of these, we generated tailored fragment spaces from sets of known inhibitors and additionally derived ranges for several physicochemical properties. We characterized the target-specific fragment spaces and were able to enumerate the complete chemical subspaces for most of the targets.

Flux (2): Comparison of Molecular Mutation and Crossover Operators for Ligand-Based de Novo Design

We implemented a fragment-based de novo design algorithm for a population-based optimization of molecular structures. The concept is grounded on an evolution strategy with mutation and crossover operators for structure breeding. Molecular building blocks were obtained from the pseudo-retrosynthesis of a collection of pharmacologically active compounds following the RECAP principle. The influence of mutation and crossover on the course of optimization was assessed in redesign studies using known drugs as template structures. A topological atom-pair descriptor grounded on potential pharmacophore points was used as a molecular descriptor, and the Manhattan distance between the template and candidate molecules served as a fitness function. Exclusive use of the crossover operator yielded few unique compounds and often resulted in premature convergence of the optimization process, whereas exclusive use of the mutation operator resulted in diverse high-quality structures. Combinations of crossover and mutation yielded the overall best results. The majority of the designed structures exhibit a chemically reasonable architecture; chiral centers are rare, and unfavorable connections of building blocks are infrequent. We conclude that this fragment-based design principle is suited as an idea generator for the automated design of novel leadlike molecules.