GeNNius: an ultrafast drug-target interaction inference method based on graph neural networks.

MOTIVATION: Drug-target interaction (DTI) prediction is a relevant but challenging task in the drug repurposing field. In-silico approaches have drawn particular attention as they can reduce associated costs and time commitment of traditional methodologies. Yet, current state-of-the-art methods present several limitations: existing DTI prediction approaches are computationally expensive, thereby hindering the ability to use large networks and exploit available datasets and, the generalization to unseen datasets of DTI prediction methods remains unexplored, which could potentially improve the development processes of DTI inferring approaches in terms of accuracy and robustness. RESULTS: In this work, we introduce GeNNius (Graph Embedding Neural Network Interaction Uncovering System), a Graph Neural Network (GNN)-based method that outperforms state-of-the-art models in terms of both accuracy and time efficiency across a variety of datasets. We also demonstrated its prediction power to uncover new interactions by evaluating not previously known DTIs for each dataset. We further assessed the generalization capability of GeNNius by training and testing it on different datasets, showing that this framework can potentially improve the DTI prediction task by training on large datasets and testing on smaller ones. Finally, we investigated qualitatively the embeddings generated by GeNNius, revealing that the GNN encoder maintains biological information after the graph convolutions while diffusing this information through nodes, eventually distinguishing protein families in the node embedding space. AVAILABILITY AND IMPLEMENTATION: GeNNius code is available at https://github.com/ubioinformat/GeNNius.

SuPepMem: A database of innate immune system peptides and their cell membrane interactions.

Host defense peptides (HDPs) are short cationic peptides that play a key role in the innate immune response of all living organisms. Their action mechanism does not depend on the presence of protein receptors, but on their ability to target and disrupt the membranes of a wide range of pathogenic and pathologic cells which are recognized by their specific compositions, typically with a relatively high concentration of anionic lipids. Lipid profile singularities have been found in cancer, inflammation, bacteria, viral infections, and even in senescent cells, enabling the possibility to use them as therapeutic targets and/or diagnostic biomarkers. Molecular dynamics (MD) simulations are extraordinarily well suited to explore how HDPs interact with membrane models, providing a large amount of qualitative and quantitative information that, nowadays, cannot be assessed by wet-lab methods at the same level of temporal and spatial resolution. Here, we present SuPepMem, an open-access repository containing MD simulations of different natural and artificial peptides with potential membrane lysis activity, interacting with membrane models of healthy mammal, bacteria, viruses, cancer or senescent cells. In addition to a description of the HDPs and the target systems, SuPepMem provides both the input files necessary to run the simulations and also the results of some selected analyses, including structural and MD-based quantitative descriptors. These descriptors are expected to be useful to train machine learning algorithms that could contribute to design new therapeutic peptides. Tools for comparative analysis between different HDPs and model membranes, as well as to restrict the queries to structural and time-averaged properties are also available. SuPepMem is a living project, that will continuously grow with more simulations including peptides of different sequences, MD simulations with different number of peptide units, more membrane models and also several resolution levels. The database is open to MD simulations from other users (after quality check by the SuPepMem team). SuPepMem is freely available under https://supepmem.com/.

The Lord of the NanoRings: Cyclodextrins and the battle against SARS-CoV-2.

A handful of singular structures and laws can be observed in nature. They are not always evident but, once discovered, it seems obvious how to take advantage of them. In chemistry, the discovery of reproducible patterns stimulates the imagination to develop new functional materials and technological or medical applications. Two clear examples are helical structures at different levels in biological polymers as well as ring and spherical structures of different size and composition. Rings are intuitively observed as holes able to thread elongated structures. A large number of real and fictional stories have rings as inanimate protagonists. The design, development or just discovering of a special ring has often been taken as a symbol of power or success. Several examples are the Piscatory Ring wore by the Pope of the Catholic Church, the NBA Championship ring and the One Ring created by the Dark Lord Sauron in the epic story The Lord of the Rings. In this work, we reveal the power of another extremely powerful kind of rings to fight against the pandemic which is currently affecting the whole world. These rings are as small as ~1 nm of diameter and so versatile that they are able to participate in the attack of viruses, and specifically SARS-CoV-2, in a large range of different ways. This includes the encapsulation and transport of specific drugs, as adjuvants to stabilize proteins, vaccines or other molecules involved in the infection, as cholesterol trappers to destabilize the virus envelope, as carriers for RNA therapies, as direct antiviral drugs and even to rescue blood coagulation upon heparin treatment. "One ring to rule them all. One ring to find them. One ring to bring them all and in the darkness bind them." J. R. R. Tolkien.