Fragment Libraries Designed to Be Functionally Diverse Recover Protein Binding Information More Efficiently Than Standard Structurally Diverse Libraries

Carbery, A.; Skyner, R.; Delft, Frank von; Deane, Charlotte M.

doi:10.1021/acs.jmedchem.2c01004

Cited by 22 publications

(28 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The diversity of chemical structures, however, does not presage the ability to provide hits for different proteins, as structurally dissimilar fragments may exhibit the same biological activity. XChem's analysis of DSiP screening campaigns thus demonstrated that structurally dissimilar fragments can form the same interactions, consistent with the pharmacophore concept (Carbery et al, 2022). The investigation of 309 proteinfragment structures obtained for 10 unrelated targets and 225 fragments led the authors to propose a strategy for selecting functionally diverse fragments based on the interactions formed with multiple proteins, as encoded in interaction fingerprints.…”

Section: Introductionmentioning

confidence: 62%

“…A minimal molecular weight around 175 Da was suggested by the study of the conservation of the binding mode of the fragments in their drug-like superstructures (Jacquemard and Kellenberger, 2019). The same threshold was proposed following the analysis of XChem's screening campaigns which revealed that fragments that have never been shown to bind to a target tend to have a low molecular weight compared to fragments that bind to one or more targets (Carbery et al, 2022). According to the size criterion, about half of the versatile PDB fragments may provide a valid base for fragment growing (105 fragments with MW > 175 Da).…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Mining the Protein Data Bank to inspire fragment library design

Imbernon

Chiesa²,

Kellenberger³

2023

Front. Chem.

View full text Add to dashboard Cite

The fragment approach has emerged as a method of choice for drug design, as it allows difficult therapeutic targets to be addressed. Success lies in the choice of the screened chemical library and the biophysical screening method, and also in the quality of the selected fragment and structural information used to develop a drug-like ligand. It has recently been proposed that promiscuous compounds, i.e., those that bind to several proteins, present an advantage for the fragment approach because they are likely to give frequent hits in screening. In this study, we searched the Protein Data Bank for fragments with multiple binding modes and targeting different sites. We identified 203 fragments represented by 90 scaffolds, some of which are not or hardly present in commercial fragment libraries. By contrast to other available fragment libraries, the studied set is enriched in fragments with a marked three-dimensional character (download at 10.5281/zenodo.7554649).

show abstract

Section: Introductionmentioning

confidence: 62%

Section: Resultsmentioning

confidence: 99%

Mining the Protein Data Bank to inspire fragment library design

Imbernon

Chiesa²,

Kellenberger³

2023

Front. Chem.

View full text Add to dashboard Cite

show abstract

“…This is a particularly useful finding given that recent research into fragment screen libraries has highlighted that even screens designed to be diverse structurally offer a limited exploration of protein pockets. 43…”

Section: Resultsmentioning

confidence: 99%

A Step Towards Generalisability: Training a Machine Learning Scoring Function for Structure-Based Virtual Screening

Scantlebury

Vost

Carbery

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Over the last few years, many new machine learning-based scoring functions for predicting the binding of small molecules to proteins have been developed. Their objective is to approximate the distribution which takes two molecules as input and outputs the energy of their interaction. This distribution is dependent on the interatomic interactions involved in binding, and only a scoring function that accounts for these interactions can accurately predict binding affinity on unseen molecules. To try to create a method capable of learning these interactions, we built PointVS: a machine learning-based scoring function which achieves state of the art performance even after rigorous filtering of the training set. This filtering is key, as we found that a commonly used benchmark, CASF-16, overestimates the true accuracy of machine learning-based scoring functions when trained using the most commonly used training set. Ranking algorithms using this benchmark rewards memorisation of training data rather than knowledge of the rules of intermolecular binding. We demonstrate that PointVS is able to identify important interactions using attribution, and further, that it can be used to extract important binding pharmacophores from a given protein target when supplied with a number of bound structures. We use this information to perform fragment elaboration, and see improvements in docking scores compared to using structural information from a traditional data-based approach. This not only provides definitive proof that PointVS is learning to identify important binding interactions, but also constitutes the first deep learning-based method for extracting structural information from a target for molecule design.

show abstract

“…The predicted interactions were compared with those of the parent fragments to discern whether the merge is able to replicate unique interactions made by each of the fragments (representing useful merges). In addition to comparing chemical diversity of the compound sets, we also compared functional diversity by looking at the interactions made by each compound set; the analysis adapts the methodology described in [57]. Interaction diversity was analysed for each target by selecting the minimal subset of filtered compounds that represent all possible interactions (compounds are ranked according to the amount of ‘information’ recovered and are added until reaching a final set in which all possible interactions are recovered).…”

Section: Methodsmentioning

confidence: 99%

The use of a graph database is a complementary approach to a classical similarity search for identifying commercially available fragment merges

Wills

Sánchez-García

Roughley

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Fragment screening using X-ray crystallography can yield rich structural data to help guide the optimization of low-molecular-weight compounds into more potent binders. Fragment merging, whereby substructural motifs from partially overlapping fragments are incorporated into a single larger compound, represents a potentially powerful and efficient approach for increasing potency. Searching commercial catalogues provides one useful way to quickly and cheaply identify follow-up compounds for purchase and further screening, and circumvents the challenge of synthetic accessibility. The Fragment Network is a graph database that provides a novel way to explore the chemical space surrounding fragment hits. We use an iteration of the database containing >120 million catalogue compounds to find fragment merges for four XChem fragment screening campaigns. Retrieved molecules were filtered using a pipeline of 2D and 3D filters and contrasted against a traditional fingerprint-based similarity search. The two search techniques were found to have complementary results, identifying merges in different regions of chemical space. Both techniques were able to identify merges that are predicted to replicate the interactions made by the parent fragments. This work demonstrates the use of the Fragment Network to increase the yield of fragment merges beyond that of a classical catalogue search, thus increasing the likelihood of finding promising follow-up compounds. We present a pipeline that is able to systematically exploit all known fragment hits by performing large-scale enumeration of all possible fragment pairs for merging.

show abstract

Fragment Libraries Designed to Be Functionally Diverse Recover Protein Binding Information More Efficiently Than Standard Structurally Diverse Libraries

Cited by 22 publications

References 34 publications

Mining the Protein Data Bank to inspire fragment library design

Mining the Protein Data Bank to inspire fragment library design

A Step Towards Generalisability: Training a Machine Learning Scoring Function for Structure-Based Virtual Screening

The use of a graph database is a complementary approach to a classical similarity search for identifying commercially available fragment merges

Contact Info

Product

Resources

About