Network analysis reveals distinct behaviors in cocrystals and solvates, implying different prediction methods are needed.
Cocrystals and solvates are two kinds of multicomponent crystals. Cocrystals consist of solid compounds, while solvates consist of a mixture of solid and liquid compounds. This work addresses the question of whether cocrystals and solvates can be treated the same or need to be treated separately. To do this, networks out of known cocrystals and solvates are created, where each node represents an individual compound, and two nodes are linked if the compounds they represent form a cocrystal or solvate. These networks are also used to predict new cocrystals and solvates using a technique called link prediction. By analyzing the structure of the cocrystal and solvate networks, differences in behaviors between cocrystals and solvates are found, due in part to a clash between chemical and steric complementarity found in the solvate network. Based on the analysis of the networks, it is found that cocrystals and solvates cannot be treated equally and that information from one network is not relevant for predictions in the other. It is also found that link prediction performs poorly when used on the solvate network, unless 14 problematic solvents are removed.
No takes yet. Share an insight, caveat, or question.
Vries et al. (2025) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: