Software toolkits for in silico screening of polymer excipients in drug formulations

Pharmaceutical formulation is a key step of the drug development process where the active
pharmaceutical ingredient (API) is mixed with non-drug components (excipients) which degrade in
the body, usually as a function of time or pH, releasing the API in the target tissue.
Biocompatible polymers like cellulose derivatives or glycols are common excipients in tablet
formulations. Polymer degradation, crystallinity and mechanical properties can be easily tuned by
changing their molecular weight or by chemical functionalisation. These features in turn affect
product shelf life, drug release time and effective API concentration in the gastrointestinal tract.
Copolymers of PEG or PVA are particularly suitable to promote the uptake of poorly soluble
drugs. However, their complex semicrystalline morphology varies with subtle changes in
composition and processing conditions, making their performance and API compatibility hard to
predict.
Selecting polymer excipients for drug formulations is a time-consuming, mostly trial-and-error
process. While in silico screening (cheminformatics, machine learning) is routinely employed in
small molecule drug discovery, such tools are not readily applicable to polymers. Simulations of
excipients are done using 'computational microscope' molecular dynamics (MD) techniques with
off-the-shelf parameters, which cannot be tailored to specific excipients. This is due to several
factors, namely the lack of automated parameterisation software - an issue Open Force Field, a
large industry backed consortium, is starting to tackle.
Accurately describing polymer flexibility is key, as torsional degrees of freedom are directly related
to aggregation and crystallinity. Overall, the lack of reliable workflows and tools for polymer
screening of functional is an open challenge across the fields of biochemistry, formulation
science, materials science.