Skip to main content

ADME SARfari: A tool for predicting and comparing cross-species ADME targets

ADME studies are focused on understanding the disposition of a compound within an organism and the results of such studies play a critical role in the drug development process. ADME studies (more commonly referred to as pharmacokinetic or PK studies) are focused on 4 main areas: Absorption, Distribution, Metabolism and Excretion. More information on the PK measurement types can be found here.

Comparisons of PK data across species is a potential problem drug researchers need to deal with, as model organism studies are the primary source of such data. For example, in an animal model study, which may be carried out on a compound as it passes through the drug development pipeline, is it meaningful to compare clearance or bioavailability data from a mouse or rat to human? Clearly there are many differences (physical, metabolic, genetic,..), which make answering these types of questions difficult. Building tools which guide researchers to potential answers or provide a better understanding of the inter-species differences are of great value - leading us nicely to the focus of this blog post.

It turns out the ChEMBL database has a wealth of PK measurements data, which allows users to start asking ADME focused questions. You can access all of this data via the ChEMBL Interface or from one of our downloads, but in order to answer some of the more complex questions a significant amount of data processing first needs to take place. So to help the ChEMBL community get started analysing this data, we, in collaboration with our colleagues at GSK, set about building a new ADME focused system. The new system is called ADME SARfari and it aims to centralise all ADME data currently stored in ChEMBL and other related databases, as well as providing new tools to help interrogate the data.

In order to build the system we have pulled data from a number of sources, which include:
  • ChEMBL - Bioactivity data, PK data and molecules.
  • PharmaADME - An online resource providing a list ADME related human genes. We used this to build our primary list of ADME Targets, but have also added a couple of extra ones.
  • The Human Protein Atlas -  Protein expression data for human ADME Targets.
  • ENSEMBL - Orthologue (using the Compara Service) and SNP data for ADME Targets.
  • The Göttingen minipig and beagle dog genome predicted ADME Targets. GSK have sequenced the  genomes of these two pharmaceutically significant species. More details on the genomes can be found here.
When you visit the site you will find it is divided into the following 7 sections:
  • Home - Allows user to initiate a compound or protein focused search.
  • Orthologues - A table of ADME orthologues. The first column of the table corresponds to the human ADME targets and additional columns correspond to targets found in model organisms.
  • Tissues - Protein expression data for human ADME targets.
  • Bioactivities - Bioactivity data and PK measurements for all ADME related targets in the system, which are also found in the ChEMBL database.
  • Molecules - Distinct set of ChEMBL molecules linked to ADME related targets via the bioactivity data.
  • Pharmacokinetic - A cross species comparison of PK data for compounds found in the  ChEMBL database  (see red heatmap image at top).
  • About - More details on how to use the system
By clicking on the links above, you can view and download all of the data associated within each of the sections. One exception is the Pharmacokinetics section, as there is too much data to display in the heatmap by default, so you must first initiate a search (we will come back to this later). Now that you have a better idea of what is in the system, what are the types of questions you can ask? Looking at the homepage you will see there are 2 ways to initiate a search of the system: Compound-initiated, using the chemical sketcher box and protein-initiated, using the BLAST or keyword search.

Protein Search  (Using BLAST)

To run a BLAST search paste a sequence in the text box to the right on the homepage:

BLAST search results will be displayed on the Orthologues page and only rows which contain hits to your search query will be returned:

Clicking on the Tissues tab displays the protein expression levels of the human targets returned by the BLAST search:

Clicking on the Bioactivities tab returns all of the ChEMBL bioactivity data for targets returned by BLAST search:

Clicking on the Molecules tab returns the distinct set of compounds currently displayed in the Bioactivities section:

Clicking on the Pharmacokinetics tab will provide a cross-species overview of the PK data in ChEMBL for the compounds currently displayed in the Molecules section (Note that not all compounds will have PK measurements):

The following points help explain what the heatmap above is displaying:
  • Each narrow row corresponds to a compound.
  • Each column corresponds to a PK measurement in a specific organism. 
  • The PK measurements summarised in this view are Clearance (Cl), Cmax, Bioavailability (f), T1/2, Tmax and Volume of Distribution (Vd). 
  • The colour of each cell corresponds to a low, medium and high binned PK measurement, making it easier to compare values across species.
  • The columns can be sorted by clicking on the header (see the second column, which corresponds to the Human Clearance data).

Compound Search

It is also possible to search the system using a compound structure. Simply draw or paste a structure into the compound sketcher on the left of the homepage:


You can then choose to run a substructure and similarity search. When you run the search you will first be taken Molecules section and you can then explore the other sections, which are all connected based on this initial set of compounds.

Predictive Model Search

The system also allows a user to predict which ADME protein target a molecule will interact with. The binding data (displayed in the Bioactivities section) has been used to build a multi-category naive Bayesian  classifier. We will follow up with a more detailed post describing the model building process, but essentially the model is able to predict if a user-submitted molecule will interact with 133 of the ADME targets included in the system. The About page provides some additional data on the targets included in the model. 

To run a search against the predictive model, draw or paste a structure into the compound sketcher on the left of the homepage and hit the "Model Prediction" button. You will be taken to the Orthologues section, but only the rows which include a target predicted to interact with submitted molecule (coloured green) will be on display:

We hope you find the ADME SARfari system useful and if you have any questions please let us know.

The ChEMBL Team


Popular posts from this blog

UniChem 2.0

UniChem new beta interface and web services We are excited to announce that our UniChem beta site will become the default one on the 11th of May. The new system will allow us to better maintain UniChem and to bring new functionality in a more sustainable way. The current interface and web services will still be reachable for a period of time at . In addition to it, the most popular legacy REST endpoints will also remain implemented in the new web services: Some downtime is expected during the swap.  What's new? UniChem’s current API and web application is implemented with a framework version that’s not maintained and the cost of updating it surpasses the cost of rebuilding it. In order to improve stability, security, and support the implementation and fast delivery of new features, we have decided to revamp our user-facing systems using the latest version of widely used and maintained frameworks, i

A python client for accessing ChEMBL web services

Motivation The CheMBL Web Services provide simple reliable programmatic access to the data stored in ChEMBL database. RESTful API approaches are quite easy to master in most languages but still require writing a few lines of code. Additionally, it can be a challenging task to write a nontrivial application using REST without any examples. These factors were the motivation for us to write a small client library for accessing web services from Python. Why Python? We choose this language because Python has become extremely popular (and still growing in use) in scientific applications; there are several Open Source chemical toolkits available in this language, and so the wealth of ChEMBL resources and functionality of those toolkits can be easily combined. Moreover, Python is a very web-friendly language and we wanted to show how easy complex resource acquisition can be expressed in Python. Reinventing the wheel? There are already some libraries providing access to ChEMBL d

LSH-based similarity search in MongoDB is faster than postgres cartridge.

TL;DR: In his excellent blog post , Matt Swain described the implementation of compound similarity searches in MongoDB . Unfortunately, Matt's approach had suboptimal ( polynomial ) time complexity with respect to decreasing similarity thresholds, which renders unsuitable for production environments. In this article, we improve on the method by enhancing it with Locality Sensitive Hashing algorithm, which significantly reduces query time and outperforms RDKit PostgreSQL cartridge . myChEMBL 21 - NoSQL edition    Given that NoSQL technologies applied to computational chemistry and cheminformatics are gaining traction and popularity, we decided to include a taster in future myChEMBL releases. Two especially appealing technologies are Neo4j and MongoDB . The former is a graph database and the latter is a BSON document storage. We would like to provide IPython notebook -based tutorials explaining how to use this software to deal with common cheminformatics p

ChEMBL 30 released

  We are pleased to announce the release of ChEMBL 30. This version of the database, prepared on 22/02/2022 contains: 2,786,911 compound records 2,157,379 compounds (of which 2,136,187 have mol files) 19,286,751 activities 1,458,215 assays 14,855 targets 84,092 documents Data can be downloaded from the ChEMBL FTP site: Please see ChEMBL_30 release notes for full details of all changes in this release: New Deposited Datasets EUbOPEN Chemogenomic Library (src_id = 55, ChEMBL Document ID CHEMBL4689842):   The EUbOPEN consortium is an Innovative Medicines Initiative (IMI) funded project to enable and unlock biology in the open. The aims of the project are to assemble an open access chemogenomic library comprising about 5,000 well annotated compounds covering roughly 1,000 different proteins, to synthesize at least

Multi-task neural network on ChEMBL with PyTorch 1.0 and RDKit

  Update: KNIME protocol with the model available thanks to Greg Landrum. Update: New code to train the model and ONNX exported trained models available in github . The use and application of multi-task neural networks is growing rapidly in cheminformatics and drug discovery. Examples can be found in the following publications: - Deep Learning as an Opportunity in VirtualScreening - Massively Multitask Networks for Drug Discovery - Beyond the hype: deep neural networks outperform established methods using a ChEMBL bioactivity benchmark set But what is a multi-task neural network? In short, it's a kind of neural network architecture that can optimise multiple classification/regression problems at the same time while taking advantage of their shared description. This blogpost gives a great overview of their architecture. All networks in references above implement the hard parameter sharing approach. So, having a set of activities relating targets and molecules we can tra