Cocrystal design by network-based link prediction

Devogelaer, Jan-Joris; Vlieg, E.; Meekes, Hugo; Tinnemans, Paul; Gelder, René de

doi:10.1039/c9ce01110b

Cited by 48 publications

(77 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, we searched for ac rystal form of 1 that does showconglomerate behavior.R ecently,o ur group has reported as ystematic approach for predicting suitable coformers to obtain stable cocrystals, some of whichp otentially being conglomerates. [26] However,g iven the complexity that such coformers might add to the system, ac onvenient ands impler route to explore conglomerate formation was to evaluate al ibrary of different sol-vents that may give rise to possible conglomerate solvates. Thus far,o nly one successful Viedma ripening deracemization involving as olvate had previously been reported, although that wasarare case of ac onglomerate co-crystal salt solvate.…”

Section: Conglomerate Screeningmentioning

confidence: 99%

Photoracemization‐Based Viedma Ripening of a BINOL Derivative

Belletti

Tortora

Mellema³

et al. 2019

Chemistry A European J

Self Cite

View full text Add to dashboard Cite

Viedma ripening is ad eracemizationp rocess that has been used to deracemize ar ange of chiral molecules. The method hast wo major requirements:t he compound needs to crystallize as ac onglomerate and it needs to be racemizable under the crystallization conditions. Although conglomerate formation can be induced in different ways, the number of racemization methods is still ratherl imited. To extend the scope of Viedma ripening, in the present research we applied UV-light-induced racemization in a Viedmar ipening process, and report the successful deracemizationo faB INOL derivative crystallizing as ac onglomerate. Irradiation by UV light activatest he target compound in combination with an organic base, required to promote the excited-state proton transfer (ESPT), leadingt hereaftert o racemization. This offers an ew tool towards the development of Viedma ripening processes, by using ac heap and "green"c atalytic source like UV light to racemize suitable chiral compounds.

show abstract

Section: Conglomerate Screeningmentioning

confidence: 99%

Photoracemization‐Based Viedma Ripening of a BINOL Derivative

Belletti

Tortora

Mellema³

et al. 2019

Chemistry A European J

Self Cite

View full text Add to dashboard Cite

show abstract

“…Therefore, it is impossible to directly extract an evidence-based list of invalid coformer combinations from the CSD, nor is screening literature for invalid combinations a feasible and unbiased option. To cope with this issue, we generated an equally large invalid cocrystal set using our link-prediction method [26], assigning a statistical likelihood to the existence of a cocrystal for two coformers based on network science. By restricting the sampled invalid cocrystals to highly unlikely combinations of coformers that each have at least 5 cocrystals in the CSD, a balanced invalid cocrystal set is found, showing a Supplementary Table 2).…”

Section: Overview Of Model Design and Selectionmentioning

confidence: 99%

Cocrystal Prediction by Artificial Neural Networks

Devogelaer¹,

Meekes²,

Tinnemans³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

<div>A significant amount of attention has been given to the design and synthesis of cocrystals by both industry and academia because of its potential to change a molecule’s physicochemical properties. This paper reports on the application of a data-driven cocrystal prediction method, based on two types of artificial neural network models and cocrystal data present in the Cambridge Structural Database. The models accept pairs of coformers and predict whether a cocrystal is likely to form.</div>

show abstract

“…Recently,w ei ntroduced ah olistic approach to study cocrystallization using network science and link prediction. [44,45] Analysis of an etwork of coformers extracted from the Cambridge Structural database (CSD) [46] shows that, rather than being ar andom assembly of coformers,i tr epresents ar ational source of co-crystal information that can form abasis for prediction. Therefore,itwould be very appealing to develop amethod that utilizes all this co-crystal information, and is able to predict co-crystals for coformers lacking any experimental data on co-crystal formation.…”

Section: Introductionmentioning

confidence: 99%

Co‐crystal Prediction by Artificial Neural Networks**

Devogelaer¹,

Meekes²,

Tinnemans³

et al. 2020

Angew Chem Int Ed

Self Cite

View full text Add to dashboard Cite

As ignificant amount of attention has been given to the design and synthesis of co-crystals by both industry and academia because of its potential to change am olecules physicochemical properties.Y et, difficulties arise when searching for adequate combinations of molecules (or coformers) to form co-crystals,h ampering the efficient exploration of the targetss olid-state landscape.T his paper reports on the application of ad ata-driven co-crystal prediction method based on two types of artificial neural network models and cocrystal data present in the Cambridge Structural Database.The models accept pairs of coformers and predict whether ac ocrystal is likely to form. By combining the output of multiple models of both types,o ur approach shows to have excellent performance on the proposed co-crystal training and validation sets,and has an estimated accuracy of 80 %for molecules for which previous co-crystallization data is unavailable.

show abstract

Cocrystal design by network-based link prediction

Abstract: Cocrystals are predicted using a network of coformers extracted from the CSD.

Cited by 48 publications

References 44 publications

Photoracemization‐Based Viedma Ripening of a BINOL Derivative

Photoracemization‐Based Viedma Ripening of a BINOL Derivative

Cocrystal Prediction by Artificial Neural Networks

Co‐crystal Prediction by Artificial Neural Networks**

Contact Info

Product

Resources

About