AllChem: generating and searching 1020 synthetically accessible structures

Target-specific novel molecules with their recipe: Incorporating synthesizability in the design process

Application of Artificial intelligence (AI) in drug discovery has led to several success stories in recent times. While traditional methods mostly relied upon screening large chemical libraries for early-stage drug-design, de novo design can help identify novel target-specific molecules by sampling from a much larger chemical space. Although this has increased the possibility of finding diverse and novel molecules from previously unexplored chemical space, this has also posed a great challenge for medicinal chemists to synthesize at least some of the de novo designed novel molecules for experimental validation. To address this challenge, in this work, we propose a novel forward synthesis-based generative AI method, which is used to explore the synthesizable chemical space. The method uses a structure-based drug design framework, where the target protein structure and a target-specific seed fragment from co-crystal structures can be the initial inputs. A random fragment from a purchasable fragment library can also be the input if a target-specific fragment is unavailable. Then a template-based forward synthesis route prediction and molecule generation is performed in parallel using the Monte Carlo Tree Search (MCTS) method where, the subsequent fragments for molecule growth can again be obtained from a purchasable fragment library. The rewards for each iteration of MCTS are computed using a drug-target affinity (DTA) model based on the docking pose of the generated reaction intermediates at the binding site of the target protein of interest. With the help of the proposed method, it is now possible to overcome one of the major obstacles posed to the AI-based drug design approaches through the ability of the method to design novel target-specific synthesizable molecules.

FragGrow: A Web Server for Structure-Based Drug Design by Fragment Growing within Constraints

Fragment growing is an important ligand design strategy in drug discovery. In this study, we present FragGrow, a web server that facilitates structure-based drug design by fragment growing. FragGrow offers two working modes: one for growing molecules through the direct replacement of hydrogen atoms or substructures and the other for growing via virtual synthesis. FragGrow works by searching for suitable fragments that meet a set of constraints from an indexed 3D fragment database and using them to create new compounds in 3D space. The users can set a range of constraints when searching for their desired fragment, including the fragment's ability to interact with specific protein sites; its size, topology, and physicochemical properties; and the presence of particular heteroatoms and functional groups within the fragment. We hope that FragGrow will serve as a useful tool for medicinal chemists in ligand design. The FragGrow server is freely available to researchers and can be accessed at https://fraggrow.xundrug.cn.

Three-dimensional quantitative structural-activity relationship and molecular dynamics study of multivariate substituted 4-oxyquinazoline HDAC6 inhibitors

3D-QSAR models were established by collecting 46 multivariate-substituted 4-oxyquinazoline HDAC6 inhibitors. The relationship of molecular structure and inhibitory activity was studied by comparative molecular field analysis (CoMFA) and comparative molecular similarity index analysis (CoMSIA). The results showed the models established by CoMFA (q² = 0.590, r² = 0.965) and CoMSIA (q² = 0.594, r² = 0.931) had good prediction ability. At the same time, 3D-QSAR models met the internal verification, external verification and AD test. Ten new compounds were designed based on CoMFA and CoMSIA contour maps and their pharmacokinetic/toxic properties (ADME/T) were evaluated. It was found that most compounds have well safety profile and pharmacokinetic property. Then, we explored the interaction between HDAC6 and compounds by molecular docking. The results showed that the binding mode of the new compounds with HDAC6 was the same as the template compound 46, and the hydrogen bond and hydrophobic bond played a vital role in the binding process. Molecular dynamics simulation results showed that residues Ser531, His574 and Tyr745 played key roles in the binding process. All newly designed compounds had lower energy gap and binding energy than compound 46 according to DFT analysis and free energy analysis. This study provided a theoretical reference for designing compounds of higher activity and a new idea for the development of novel HDAC6 inhibitors.

SynthI: A New Open-Source Tool for Synthon-Based Library Design

Most of the existing computational tools for de novo library design are focused on the generation, rational selection, and combination of promising structural motifs to form members of the new library. However, the absence of a direct link between the chemical space of the retrosynthetically generated fragments and the pool of available reagents makes such approaches appear as rather theoretical and reality-disconnected. In this context, here we present Synthons Interpreter (SynthI), a new open-source toolkit for de novo library design that allows merging those two chemical spaces into a single synthons space. Here synthons are defined as actual fragments with valid valences and special labels, specifying the position and the nature of reactive centers. They can be issued from either the "breakup" of reference compounds according to 38 retrosynthetic rules or real reagents, after leaving group withdrawal or transformation. Such an approach not only enables the design of synthetically accessible libraries and analog generation but also facilitates reagents (building blocks) analysis in the medicinal chemistry context. SynthI code is publicly available at https://github.com/Laboratoire-de-Chemoinformatique/SynthI.

Topological Similarity Search in Large Combinatorial Fragment Spaces

In similarity-driven virtual screening, molecular fingerprints are widely used to assess the similarity of all compounds contained in a chemical library to a query compound of interest. This similarity analysis is traditionally done for each member of the library individually. When encoding chemical spaces that surpass billions of compounds in size, it becomes impractical to enumerate all their products, let alone assess their similarity, deeming this approach impossible without investing a substantial amount of resources. In this work, we present a novel search algorithm named SpaceLight for topological fingerprint similarity searching in large, practically non-enumerable combinatorial fragment spaces. In contrast to existing methods, SpaceLight is able to utilize the combinatorial character of these chemical spaces for efficiency while maintaining a high correlation of the description of molecular similarity to well-known molecular fingerprints like ECFP. The resulting software is able to search prominent spaces like EnamineREAL with more than 10 billion compounds in seconds on a standard desktop computer.

Comparison of Large Chemical Spaces

Chemical libraries are commonplace in computer-aided drug discovery, and assessing their overlap/complementarity is a routine task. For this purpose, different techniques are applied, ranging from exact matching to comparing physicochemical properties. However, these techniques are applicable only if the compound sets are not too big. Particularly for chemical spaces, containing billions of compounds, alternative ways of assessment are required. Random subsets could be enumerated and compared one-to-one, but given the vast sizes of the chemical spaces assessed here, such samples can at best provide a rough estimate of any overlap. Here we describe a novel way to compare chemical spaces utilizing a panel of query compounds. We applied this technique to three different types of spaces and obtained insight into their structural overlap, their coverage of the chemical universe, and their density. As chemical feasibility of virtual compounds is particularly important, we included related in silico predictions in our assessment.

Virtual Compound Libraries in Computer-Assisted Drug Discovery

The use of virtual compound libraries in computer-assisted drug discovery has gained in popularity and has already lead to numerous successes. Here, we examine key static and dynamic virtual library concepts that have been developed over the past decade. To facilitate the search for new drugs in the vastness of chemical space, there are still several hurdles to overcome, including the current difficulties in screening and parsing efficiency and the need for more reliable vendors and accurate synthesis prediction tools. These challenges should be tackled by both the developers of virtual libraries and by their users, in order for the exploration of chemical space to live up to its potential.

Compound property enhancement by virtual compound synthesis

During drug discovery, drug candidates are narrowed down over several steps to develop pharmaceutical products. The theoretical chemical space in such steps is estimated to be [Formula: see text]. To cover that space, extensive virtual compound libraries have been developed; however, the compilation of extensive libraries comes at large computational cost. Thus, to reduce the computational cost, researchers have constructed custom-made virtual compound libraries that focus on target diseases. In this study, we develop a system that generates virtual compound libraries from input compounds. When a user inputs a compound, the system recursively applies virtual synthetic reaction rules to the compound to improve its properties. The synthetic pathway can also be traced by the user because the reaction rules in this system are based on real organic synthesis reactions. This system has useful functions for effective drug design, such as structural preservation, allowing the substructures necessary for potency to be maintained. In this paper, to confirm the effect of directional reaction sets, we applied the reaction sets to 100 compounds. Moreover, to confirm that the system can reproduce real synthetic pathways, the synthetic pathways of Ibuprofen and Ofloxacin were explored by inputting isobutyl benzene and 7,8-difluoro-2,3-dihydro-3-methyl-4H-benzoxazine. This application is available at the following URL: http://enh.sekijima-lab.org .

A new paradigm for designing ring construction strategies for green organic synthesis: implications for the discovery of multicomponent reactions to build molecules containing a single ring

A new way of developing novel synthesis strategies for the construction of monocyclic rings found in organic molecules is presented. The method is based on the visual application of integer partitioning to chemical structures. Two problems are addressed: (1) the determination of the total number of possible ways to construct a given ring by 2-, 3-, and 4-component couplings; and (2) the systematic enumeration of those possibilities. The results of the method are illustrated using cyclohexanone, pyrazole, and the Biginelli adduct as target ring systems with a view to discover new and greener strategies for their construction using multicomponent reactions. The application of the method is also extended to various heterocycles found in many natural products and pharmaceuticals.

Multistep Reaction Based De Novo Drug Design: Generating Synthetically Feasible Design Ideas

We describe a "multistep reaction driven" evolutionary algorithm approach to de novo molecular design. Structures generated by the approach include a proposed synthesis path intended to aid the chemist in assessing the synthetic feasibility of the ideas that are generated. The methodology is independent of how the design ideas are scored, allowing multicriteria drug design to address multiple issues including activity at one or more pharmacological targets, selectivity, physical and ADME properties, and off target liabilities; the methods are compatible with common computer-aided drug discovery "scoring" methodologies such as 2D- and 3D-ligand similarity, docking, desirability functions based on physiochemical properties, and/or predictions from 2D/3D QSAR or machine learning models and combinations thereof to be used to guide design. We have performed experiments to assess the extent to which known drug space can be covered by our approach. Using a library of 88 generic reactions and a database of ∼20 000 reactants, we find that our methods can identify "close" analogs for ∼50% of the known small molecule drugs with molecular weight less than 300. To assess the quality of the in silico generated synthetic pathways, synthesis chemists were asked to rate the viability of synthesis pathways: both "real" and in silico generated. In silico reaction schemes generated by our methods were rated as very plausible with scores similar to known literature synthesis schemes.

SCUBIDOO: A Large yet Screenable and Easily Searchable Database of Computationally Created Chemical Compounds Optimized toward High Likelihood of Synthetic Tractability

De novo drug design is widely assisted by computational approaches that enable the generation of a tremendous amount of new virtual molecules within a short time frame. While the novelty of the computationally generated compounds can easily be assessed, such approaches often neglect the synthetic feasibility of the molecules, thus creating a potential hurdle that can be a barrier to further investigation. Therefore, we have developed SCUBIDOO, a freely accessible database concept that currently holds 21 million virtual products originating from a small library of building blocks and a collection of robust organic reactions. This large data set was reduced to three representative and computationally tractable samples denoted as S, M, and L, containing 9994, 99,977, and 999,794 products, respectively. These small sets are useful as starting points for ligand identification and optimization projects. The generated products come with synthesis instructions and alerts of possible side reactions, and we show that they exhibit drug-like properties while still extending into unexplored quadrants of chemical space, thus suggesting novelty. We show multiple examples that demonstrate how SCUBIDOO can facilitate the search around initial hits. This database might be a useful idea generator for early ligand discovery projects since it allows a focus on those molecules that are likely to be synthetically feasible and can therefore be studied further. Together with its modular building block construction principle, this database is also suitable for structure-activity relationship studies or fragment-growing strategies.

Computational methods for fragment-based ligand design: growing and linking

Fragment-based drug design has proved itself as a powerful technique for increasing the sampling and diversity of chemical space and enabling the design of novel leads and compounds. Computational techniques for identifying fragments, binding sites and particularly for linking, growing, and evolving fragments play a significant role in the process. Information from ADME studies and clustering property information in the form of toxicophores and chemotypes can play a significant role in aiding the design of novel, selective fragments with good activity profiles.

Template-based de novo design for type II kinase inhibitors and its extented application to acetylcholinesterase inhibitors

There is a compelling need to discover type II inhibitors targeting the unique DFG-out inactive kinase conformation since they are likely to possess greater potency and selectivity relative to traditional type I inhibitors. Using a known inhibitor, such as a currently available and approved drug or inhibitor, as a template to design new drugs via computational de novo design is helpful when working with known ligand-receptor interactions. This study proposes a new template-based de novo design protocol to discover new inhibitors that preserve and also optimize the binding interactions of the type II kinase template. First, sorafenib (Nexavar) and nilotinib (Tasigna), two type II inhibitors with different ligand-receptor interactions, were selected as the template compounds. The five-step protocol can reassemble each drug from a large fragment library. Our procedure demonstrates that the selected template compounds can be successfully reassembled while the key ligand-receptor interactions are preserved. Furthermore, to demonstrate that the algorithm is able to construct more potent compounds, we considered kinase inhibitors and other protein dataset, acetylcholinesterase (AChE) inhibitors. The de novo optimization was initiated using a template compound possessing a less than optimal activity from a series of aminoisoquinoline and TAK-285 inhibiting type II kinases, and E2020 derivatives inhibiting AChE respectively. Three compounds with greater potency than the template compound were discovered that were also included in the original congeneric series. This template-based lead optimization protocol with the fragment library can help to design compounds with preferred binding interactions of known inhibitors automatically and further optimize the compounds in the binding pockets.

Very large virtual compound spaces: construction, storage and utility in drug discovery

Recent activities in the construction, storage and exploration of very large virtual compound spaces are reviewed by this report. As expected, the systematic exploration of compound spaces at the highest resolution (individual atoms and bonds) is intrinsically intractable. By contrast, by staying within a finite number of reactions and a finite number of reactants or fragments, several virtual compound spaces have been constructed in a combinatorial fashion with sizes ranging from 10(11)11 to 10(20)20 compounds. Multiple search methods have been developed to perform searches (e.g. similarity, exact and substructure) into those compound spaces without the need for full enumeration. The up-front investment spent on synthetic feasibility during the construction of some of those virtual compound spaces enables a wider adoption by medicinal chemists to design and synthesize important compounds for drug discovery. Recent activities in the area of exploring virtual compound spaces via the evolutionary approach based on Genetic Algorithm also suggests a positive shift of focus from method development to workflow, integration and ease of use, all of which are required for this approach to be widely adopted by medicinal chemists.

Comprehensive 3D-QSAR and binding mode of BACE-1 inhibitors using R-group search and molecular docking

The β-enzyme (BACE), which takes an active part in the processing of amyloid precursor protein, thereby leads to the production of amyloid-β (Aβ) in the brain, is a major therapeutic target against Alzheimer's disease. The present study is aimed at studying 3D-QSAR of BACE-1 inhibitors and their binding mode. We build a 3D-QSAR model involving 99 training BACE-1 inhibitors based on Topomer CoMFA, and 26 molecules are employed to validate the external predictive power of the model obtained. The multiple correlation coefficients of fitting modeling, leave one out cross validation, and external validation are 0.966, 0.767 and 0.784, respectively. Topomer search is used as a tool for virtual screening in lead-like compounds of ZINC databases (2012); as a result, we successfully design 30 new molecules with higher activity than that of all training and test inhibitors. Besides, Surflex-dock is employed to explore binding mode of the inhibitors studied when acting with BACE-1 enzyme. The result shows that the inhibitors closely interact with the key sites related to ASP93, THR133, GLN134, ASP289, GLY291, THR292, THR293, ASN294, ARG296 and SER386 of BACE-1.

Pfizer Global Virtual Library (PGVL): a chemistry design tool powered by experimentally validated parallel synthesis information

An unprecedented amount of parallel synthesis information was accumulated within Pfizer over the past 12 years. This information was captured by an informatics tool known as PGVL (Pfizer Global Virtual Library). PGVL was used for many aspects of drug discovery including automated reactant mining and reaction product formation to build a synthetically feasible virtual compound collection. In this report, PGVL is discussed in detail. The chemistry information within PGVL has been used to extract synthesis and design information using an intuitive desktop Graphic User Interface, PGVL Hub. Several real-case examples of PGVL are also presented.

Mining electronic laboratory notebooks: analysis, retrosynthesis, and reaction based enumeration

An approach to automatically analyze and use the knowledge contained in electronic laboratory notebooks (ELNs) has been developed. Reactions were reduced to their reactive center and converted to a string representation (SMIRKS) which formed the basis for reaction classification and in silico (retro-)synthesis. Of the SMIRKS that occurred at least five times, 98% successfully regenerated the original product. The extracted reaction rules (SMIRKS) and corresponding reactants span a virtual chemical space which showed a strong dependence on the size of the reactive center. Whereas relatively few robust reaction types were sufficient to describe a large part of all reactions, considerably more reaction rules were necessary to cover all reactions in the ELN. Furthermore, reaction sequences were extracted to identify frequent combinations and diversifying reaction steps. Based on the extracted knowledge a (retro-)synthesis tool was built allowing for de novo design of compounds which have a high chance of being synthetically accessible. In an example application of the de novo design tool, various feasible retrosynthetic routes to the query molecule were obtained. Reaction based enumeration along the top ranked route yielded a library of 29 920 compounds with diverse properties, 99.9% of which are novel in the sense that they are unknown to the public domain.

LEAP into the Pfizer Global Virtual Library (PGVL) space: creation of readily synthesizable design ideas automatically

Pfizer Global Virtual Library (PGVL) of 10(13) readily synthesizable molecules offers a tremendous opportunity for lead optimization and scaffold hopping in drug discovery projects. However, mining into a chemical space of this size presents a challenge for the concomitant design informatics due to the fact that standard molecular similarity searches against a collection of explicit molecules cannot be utilized, since no chemical information system could create and manage more than 10(8) explicit molecules. Nevertheless, by accepting a tolerable level of false negatives in search results, we were able to bypass the need for full 10(13) enumeration and enabled the efficient similarity search and retrieval into this huge chemical space for practical usage by medicinal chemists. In this report, two search methods (LEAP1 and LEAP2) are presented. The first method uses PGVL reaction knowledge to disassemble the incoming search query molecule into a set of reactants and then uses reactant-level similarities into actual available starting materials to focus on a much smaller sub-region of the full virtual library compound space. This sub-region is then explicitly enumerated and searched via a standard similarity method using the original query molecule. The second method uses a fuzzy mapping onto candidate reactions and does not require exact disassembly of the incoming query molecule. Instead Basis Products (or capped reactants) are mapped into the query molecule and the resultant asymmetric similarity scores are used to prioritize the corresponding reactions and reactant sets. All sets of Basis Products are inherently indexed to specific reactions and specific starting materials. This again allows focusing on a much smaller sub-region for explicit enumeration and subsequent standard product-level similarity search. A set of validation studies were conducted. The results have shown that the level of false negatives for the disassembly-based method is acceptable when the query molecule can be recognized for exact disassembly, and the fuzzy reaction mapping method based on Basis Products has an even better performance in terms of lower false-negative rate because it is not limited by the requirement that the query molecule needs to be recognized by any disassembly algorithm. Both search methods have been implemented and accessed through a powerful desktop molecular design tool (see ref. (33) for details). The chapter will end with a comparison of published search methods against large virtual chemical space.

Computational medicinal chemistry in fragment-based drug discovery: what, how and when

The use of fragment-based drug discovery (FBDD) has increased in the last decade due to the encouraging results obtained to date. In this scenario, computational approaches, together with experimental information, play an important role to guide and speed up the process. By default, FBDD is generally considered as a constructive approach. However, such additive behavior is not always present, therefore, simple fragment maturation will not always deliver the expected results. In this review, computational approaches utilized in FBDD are reported together with real case studies, where applicability domains are exemplified, in order to analyze them, and then, maximize their performance and reliability. Thus, a proper use of these computational tools can minimize misleading conclusions, keeping the credit on FBDD strategy, as well as achieve higher impact in the drug-discovery process. FBDD goes one step beyond a simple constructive approach. A broad set of computational tools: docking, R group quantitative structure–activity relationship, fragmentation tools, fragments management tools, patents analysis and fragment-hopping, for example, can be utilized in FBDD, providing a clear positive impact if they are utilized in the proper scenario – what, how and when. An initial assessment of additive/non-additive behavior is a critical point to define the most convenient approach for fragments elaboration.

No electron left behind: a rule-based expert system to predict chemical reactions and reaction mechanisms

Predicting the course and major products of arbitrary reactions is a fundamental problem in chemistry, one that chemists must address in a variety of tasks ranging from synthesis design to reaction discovery. Described here is an expert system to predict organic chemical reactions based on a knowledge base of over 1500 manually composed reaction transformation rules. Novel rule extensions are introduced to enable robust predictions and describe detailed reaction mechanisms at the level of electron flows in elementary reaction steps, ensuring that all reactions are properly balanced and atom-mapped. The core reaction prediction functionalities of this expert system are illustrated with applications including: (1) prediction of detailed reaction mechanisms; (2) computer-based learning in organic chemistry; (3) retrosynthetic analysis; and (4) combinatorial library design. Select applications are available via http://cdb.ics.uci.edu.

Identification of novel non-hydroxamate anthrax toxin lethal factor inhibitors by topomeric searching, docking and scoring, and in vitro screening

Anthrax is an infectious disease caused by Bacillus anthracis, a Gram-positive, rod-shaped, anaerobic bacterium. The lethal factor (LF) enzyme is secreted by B. anthracis as part of a tripartite exotoxin and is chiefly responsible for anthrax-related cytotoxicity. As LF can remain in the system long after antibiotics have eradicated B. anthracis from the body, the preferred therapeutic modality would be the administration of antibiotics together with an effective LF inhibitor. Although LF has garnered a great deal of attention as an attractive target for rational drug design, relatively few published inhibitors have demonstrated activity in cell-based assays and, to date, no LF inhibitor is available as a therapeutic or preventive agent. Here we present a novel in silico high-throughput virtual screening protocol that successfully identified 5 non-hydroxamic acid small molecules as new, preliminary LF inhibitor scaffolds with low micromolar inhibition against that target, resulting in a 12.8% experimental hit rate. This protocol screened approximately 35 million nonredundant compounds for potential activity against LF and comprised topomeric searching, docking and scoring, and drug-like filtering. Among these 5 hit compounds, none of which has previously been identified as a LF inhibitor, three exhibited experimental IC(50) values less than 100 microM. These three preliminary hits may potentially serve as scaffolds for lead optimization as well as templates for probe compounds to be used in mechanistic studies. Notably, our docking simulations predicted that these novel hits are likely to engage in critical ligand-receptor interactions with nearby residues in at least two of the three (S1', S1-S2, and S2') subsites in the LF substrate binding area. Further experimental characterization of these compounds is in process. We found that micromolar-level LF inhibition can be attained by compounds with non-hydroxamate zinc-binding groups that exhibit monodentate zinc chelation as long as key hydrophobic interactions with at least two LF subsites are retained.

Heteroaromatic rings of the future

Small aromatic ring systems are of central importance in the development of novel synthetic protein ligands. Here we generate a complete list of 24,847 such ring systems. We call this list and associated annotations VEHICLe, which stands for virtual exploratory heterocyclic library. Searches of literature and compound databases, using this list as substructure queries, identified only 1701 as synthesized. Using a carefully validated machine learning approach, we were able to estimate that the number of unpublished, but synthetically tractable, VEHICLe rings could be over 3000. However, analysis also shows that the rate of publication of novel examples to be as low as 5-10 per year. With this work, we aim to provide fresh stimulus to creative organic chemists by highlighting a small set of apparently simple ring systems that are predicted to be tractable but are, to the best of our knowledge, unconquered.

Virtual screening for R-groups, including predicted pIC50 contributions, within large structural databases, using Topomer CoMFA

Multiple R-groups (monovalent fragments) are implicitly accessible within most of the molecular structures that populate large structural databases. R-group searching would desirably consider pIC50 contribution forecasts as well as ligand similarities or docking scores. However, R-group searching, with or without pIC50 forecasts, is currently not practical. The most prevalent and reliable source of pIC50 predictions, existing 3D-QSAR approaches, is also difficult and somewhat subjective. Yet in 25 of 25 trials on data sets on which a field-based 3D-QSAR treatment had already succeeded, substitution of objective (canonically generated) topomer poses for the original structure-guided manual alignments produced acceptable 3D-QSAR models, on average having almost equivalent statistical quality to the published models, and with negligible effort. Their overall pIC50 prediction error is 0.805, calculated as the average over these 25 topomer CoMFA models in the standard deviations of pIC50 predictions, derived from the 1109 possible "leave-out-one-R-group" (LOORG) pIC50 contributions. (This novel LOORG protocol provides a more realistic and stringent test of prediction accuracy than the customary "leave-out-one-compound" LOO approach.) The associated average predictive r(2) of 0.495 indicates a pIC50 prediction accuracy roughly halfway between perfect and useless. To assess the ability of topomer-CoMFA based virtual screening to identify "highly active" R-groups, a Receiver Operating Curve (ROC) approach was adopted. Using, as the binary criterion for a "highly active" R-group, a predicted pIC50 greater than the top 25% of the observed pIC50 range, the ROC area averaged across the 25 topomer CoMFA models is 0.729. Conventionally interpreted, the odds that a "highly active" R-group will indeed confer such a high pIC50 are 0.729/(1-0.729) or almost 3 to 1. To confirm that virtual screening within large collections of realized structures would provide a useful quantity and variety of R-group suggestions, combining shape similarity with the "highly active" pIC50, the 50 searches provided by these 25 models were applied to 2.2 million structurally distinct R-group candidates among 2.0 million structures within a ZINC database, identifying an average of 5705 R-groups per search, with the highest predicted pIC50 combination averaging 1.6 log units greater than the highest reported pIC50s.

Tagged fragment method for evolutionary structure-based de novo lead generation and optimization

Here we describe a computer-assisted de novo drug design method, EAISFD, which combines the de novo design engine EA-Inventor with a scoring function featuring the molecular docking program Surflex-Dock. This method employs tagged fragments, which are preserved substructures in EA-Inventor used for base fragment matching in Surflex-Dock, for constructing ligand structures under specific binding motifs. In addition, a target score mechanism is adopted that allows EAISFD to deliver a diverse set of desired structures. This method can be used to design novel ligand scaffolds (lead generation) or to optimize attachments on a fixed scaffold (lead optimization). EAISFD has successfully suggested many known inhibitor scaffolds as well as a number of new scaffold types when applied to p38 MAP kinase.