A strategy for the design of molecules with large two-photon absorption cross sections, delta, was developed, on the basis of the concept that symmetric charge transfer, from the ends of a conjugated system to the middle, or vice versa, upon excitation is correlated to enhanced values of delta. Synthesized bis(styryl)benzene derivatives with donor-pi-donor, donor-acceptor-donor, and acceptor-donor-acceptor structural motifs exhibit exceptionally large values of delta, up to about 400 times that of trans-stilbene. Quantum chemical calculations performed on these molecules indicate that substantial symmetric charge redistribution occurs upon excitation and provide delta values in good agreement with experimental values. The combination of large delta and high fluorescence quantum yield or triplet yield exhibited by molecules developed here offers potential for unprecedented brightness in two-photon fluorescent imaging or enhanced photosensitivity in two-photon sensitization, respectively.

In de novo drug design, computational strategies are used to generate novel molecules with good affinity to the desired biological target. In this work, we show that recurrent neural networks can be trained as generative models for molecular structures, similar to statistical language models in natural language processing. We demonstrate that the properties of the generated molecules correlate very well with the properties of the molecules used to train the model. In order to enrich libraries with molecules active toward a given biological target, we propose to fine-tune the model with small sets of molecules, which are known to be active against that target. Against Staphylococcus aureus, the model reproduced 14% of 6051 hold-out test molecules that medicinal chemists designed, whereas against Plasmodium falciparum (Malaria), it reproduced 28% of 1240 test molecules. When coupled with a scoring function, our model can perform the complete de novo drug design cycle to generate large sets of novel molecules for drug discovery.

No abstract