The KLIFS database is a rich resource for datasets focused on kinase pockets, ranging from annotated pockets and ligand interaction patterns in experimental PDB structures to ligand bioactivity values from the ChEMBL database. It is possible to explore the KLIFS resource via their web interface online (NAR 2021) or locally using dedicated KNIME nodes (JCIM 2017 and ChemMedChem 2018 developed by the KLIFS authors. With OpenCADD-KLIFS, we now add a Python solution for easy and quick integration of KLIFS datasets into Python-based pipelines.
Digidrug First up, on Thu 17. February 2022, 4pm - 5:30pm CET the next episode of the virtual seminar series, Digital Science for Drug Discovery will be held. Focused on the Berlin region, the series aims to facilitate and enmesh relationships between researchers working within both academia and industry. The over-arching theme? Making efficient and creative use of the wealth of available and growing chemical and biological data combined with powerful computational means at our disposal.
Kinases are an important class of drug targets since their dysregulation can cause severe diseases such as cancer, inflammation, and neurodegeneration. Finding selective drugs, however, is challenging due to the highly conserved binding sites across the roughly 500 human kinases, which can lead to unwanted side-effects. The underlying off-targets are often not trivial to predict or to explain from a sequence-based perspective. In our KiSSim project, we explored kinase similarity from a physicochemical and structural point of view.
Assessing kinase similarity, or how to avoid side effects? Kinases are highly conserved in their binding site, which presents a challenge since one ligand may bind not only the designated target (on-target) but also other targets (off-targets), causing mild to severe sides effects. Being able to assess kinase similarity could therefore give insight into potential side-effects. The proposed pipeline, part of the TeachOpenCADD project, allows the study of kinase similarities from four different angles in an automated and modular fashion.
Have you heard of TeachOpenCADD? Yes? Perfect, we have good news! We are back with more content (12 fresh new notebooks!), a new website, simplified installation options, and much, much more. No, you haven’t? Well, we are happy you found your way here! We will show you what TeachOpenCADD has to offer and tell you more about the exciting details on the new release!
Are you tired of hearing that deep learning needs more data and that physico-chemical data sets are still scarce? Check out our latest publication Kimber, 2021 for some augmentation strategies you can apply to improve your predictions when training deep neural networks using SMILES. You can even make predictions for novel compounds using the command-line interface. The code is open-source and available on GitHub at https://github.com/volkamerlab/maxsmi.
Our collaborative work with the BfR - combining in vitro with in silico predictions - is now published in the Environment International Journal: “Quantitative high-throughput phenotypic screening for environmental estrogens using the E-Morph Screening Assay in combination with in silico predictions” (DOI: 10.1016/j.envint.2021.106947).
Have a look at our PLIPify work (WIP)! PLIPify provides a wrapper around PLIP, which allows to digest multiple structures at once, performs the mapping of the individual profiles to fingerprints and reports protein-ligand interaction frequencies. It has recently been applied to detect the main interaction within a set of roughly 400 MPro complex structures (large crystal-based fragment screen by Diamond Light Source) in the broader scope of the COVID Moonshot initiative.
Save the date (Thu. 9. September 2021, 4pm - 5:30pm CET) for an upcoming virtual seminar series, Digital Science for Drug Discovery. Focused on the Berlin region, the series aims to facilitate and enmesh relationships between researchers working within both academia and industry. The over-arching theme? Making efficient and creative use of the wealth of available and growing chemical and biological data combined with powerful computational means at our disposal.
Interested in the new developments of deep learning in virtual screening ? Check out our review Kimber, 2021. We discuss ligand, protein and complex encodings, deep learning models, data sets, and recent studies. If you want to generate different encodings for your figures, have a look at the GitHub repository!