The ChEMBL-og

During drug development, active pharmaceutical ingredients are often formulated as salts to provide the final pharmaceutical product. ChEMBL includes parent molecules and their salts (approved and investigational) as well as other alternative forms such as hydrates and radioisotopes. These alternative forms are linked to their parent compound through the molecule hierarchy. Using the molecule hierarchyThe molecule hierarchy can be used to retrieve and display connected compounds and to aggregate activity data that has been mapped to any member of a compound family. On the interface, related compounds are automatically displayed in the ‘Alternative forms’ section of the ChEMBL compound report card. Bioactivity data can easily be aggregated in the activity summary by using the 'Include/Exclude Alternative Forms' filter.

Finding the molecule hierarchyOn the interface, we include alternative forms as shown above. The downloaded database contains the molecule_hierarchy table and the …

We’re sometimes asked what the ‘activity_comments’ in the ChEMBL database mean. In this Blog post, we’ll use aspirin as an example to explain some of the more common activity comments.

First, let’s review the bioactivity data included in ChEMBL. We extract bioactivity data directly from  seven core medicinal chemistry journals. Some common activity types, such as IC50s, are standardised to allow broad comparisons across assays; the standardised data can be found in the standard_value, standard_relation and standard_units fields. Original data is retained in the database downloads in the value, relation and units fields.



However, we extract all data from a publication including non-numerical bioactivity and ADME data. In these cases, the activity comments may be populated during the ChEMBL extraction-curation process in order to capture the author's overall conclusions.

Similarly, for deposited datasets and subsets of other databases (e.g. DrugMatrix, PubChem), the activity_comments r…

FPSim2 is the fast Python3 similarity search tool we are currently using in the ChEMBL interface. It's been a while since we first (and last) posted about it so we thought it deserved an update.

We've just released a new version (v0.2.0) and the highlights since we first talked about it are:
CPU intensive functions moved from Cython to C++ with Pybind11. Now also using libpopcnt
Improved speed, specially when dealing with some edge casesConda builds avaiable for Windows, Mac and Linux. There is no Conda for ARM but it also compiles and works in a Raspberry Pi! (and probably will do with the new ARM Macs as well)
Tversky search with a and b parameters (it previously had the 'substructure' feature with a and b respectively fixed to 1 and 0)Distance matrix calculation of the whole set feature is now availableZenodo DOI also available: 10.5281/zenodo.3902922
From a user point of view, the most interesting new feature is probably the distance matrix calculation. After the finger…

What a time! For most of us, this is the first time that we have experienced a pandemic and its impact on our daily life. Although working from home has become our routine in the ChEMBL group, we are still working as hard as ever! Of course, COVID-19 data is taking up some of our attention,  (see ChEMBL_27) but we are also continuing our work relevant to other diseases that affect large populations around the world.

Today, we are going to talk about malaria. As you may know, this disease of the Plasmodium parasite family threatens nearly half of the world’s population and led to over 400,000 deaths in 2019, predominantly among children in resource-limited areas in Africa, Asia and Central and South America. New therapies are desperately needed, in particular to cope with increased resistance against artemisinin-based combination therapies.

To help address this challenge, we have been involved in a public-private consortium that aims to deliver a tool to predict potential bl…

The COVID-19 pandemic has resulted in an unprecedented effort across the global scientific community. Drug discovery groups are contributing in several ways, including the screening of compounds to identify those with potential anti-SARS-CoV-2 activity. When the compounds being assayed are marketed drugs or compounds in clinical development then this may identify potential repurposing opportunities (though there are many other factors to consider including safety and PK/PD considerations; see for example https://www.medrxiv.org/content/10.1101/2020.04.16.20068379v1.full.pdf+html). The results from such compound screening can also help inform and drive our understanding of the complex interplay between virus and host at different stages of infection.
Several large-scale drug screening studies have now been described and made available as pre-prints or as peer-reviewed publications. The ChEMBL team has been following these developments with significant interest, and as a contribution t…

Python is nowadays the favourite platform for many machine learning scientists. It is an easy to learn language that provides a vast number of nice data science and AI tools perfect for rapid prototyping.

Models trained using Python DNN libraries like PyTorch and Tensorflow usually perform well enough to be used for production runs, but there are some situations that require the predictions to be run in C++ i.e., when best performance is required or when the model needs to be integrated with an existing C++ codebase.

ONNX (Open Neural Network Exchange) is an AI framework designed to allow interoperability between ML/DL frameworks. It allows, for example, models trained in scikit-learn, PyTorch, TensorFlow and other popular frameworks to be converted to the "standard" ONNX format for later use in any programming language with an existing ONNX runtime.

In this mini-project we trained a "dummy" DNN network (single task target predictor) in Python with PyTorch, exporte…