No abstract

In recent years, the debate in the field of applications of Deep Learning to Virtual Screening has focused on the use of neural embeddings with respect to classical descriptors in order to encode both structural and physical properties of ligands and/or targets. The attention on embeddings with the increasing use of Graph Neural Networks aimed at overcoming molecular fingerprints that are short range embeddings for atomic neighborhoods. Here, we present EMBER, a novel molecular embedding made by seven molecular fingerprints arranged as different “spectra” to describe the same molecule, and we prove its effectiveness by using deep convolutional architecture that assesses ligands’ bioactivity on a data set containing twenty protein kinases with similar binding sites to CDK1. The data set itself is presented, and the architecture is explained in detail along with its training procedure. We report experimental results and an explainability analysis to assess the contribution of each fingerprint to different targets.

Background A Virtual Screening algorithm has to adapt to the different stages of this process. Early screening needs to ensure that all bioactive compounds are ranked in the first positions despite of the number of false positives, while a second screening round is aimed at increasing the prediction accuracy. Results A novel CNN architecture is presented to this aim, which predicts bioactivity of candidate compounds on CDK1 using a combination of molecular fingerprints as their vector representation, and has been trained suitably to achieve good results as regards both enrichment factor and accuracy in different screening modes (98.55% accuracy in active-only selection, and 98.88% in high precision discrimination). Conclusion The proposed architecture outperforms state-of-the-art ML approaches, and some interesting insights on molecular fingerprints are devised.

During the last decades, biomaterials have been deeply studied to perform and improve coatings for biomedical devices. Metallic materials, especially in the orthopedic field, represent the most common material used for different type of devices thanks to their good mechanical properties. Nevertheless, low/medium resistance to corrosion and low osteointegration ability characterizes these materials. To overcome these problems, the use of biocoatings on metals substrate is largely diffused. In fact, biocoatings have a key role to confer biocompatibility properties, to inhibit corrosion and thus improve the lifetime of implanted devices. In this work, the attention was focused on Hydroxyapatite-Chitosan (HA/CS) and Hydroxyapatite-Polyvinylacetate (HA/PVAc) composites, that have been studied as biocoatings for 304 SS based devices. Hydroxyapatite was selected for its osteoconductivity thanks to its chemical structure similar to bones. Furthermore, Chitosan and Polyvinylacetate are largely used yet in medical field (e.g. antibacterial agent or drug deliver) and in this work were used to create a synergic interaction with hydroxyapatite to increase the strength and bioactivity of coating. Biocotings were obtained by galvanic deposition process that does not require an external power supply. It is a spontaneous electrochemical deposition in which materials with different standard electrochemical potential were short-circuited and immersed in an electrolytic solution. Electrons supply for the cathodic reaction in the noblest material comes from oxidation of the less noble material. SEM, EDS, XRD and RAMAN were performed for chemical-physics characterization of biocoatings. Polarization and impedance measurements have been carried out to evaluate corrosion behavior. Besides, in-vitro cytotoxicity assays have been done for the biological features.

No abstract