How much does drug discovery software cost?
How number of users affect drug discovery software costs The number of people who need access to the platform is…
The Southern Research Institute has a broad interdisciplinary program with a wealth of success in drug discovery, including 8 FDA-approved cancer drugs. We met with Dr Sixue Zhang to discuss how and why their scientists use StarDrop to guide their molecular design and data analysis.
Tell us about your recent project. What challenges did you face?
SZ: Drug discovery is a multi-parameter challenge. A drug hunter needs to consider not only activity, but pharmacokinetics and ADME properties too, to design balanced molecules.
Our recent collaboration with the University of Alabama was focused on new anti-inflammation agents, important in treating acute skin, kidney and liver injuries. Our team, consisting of computational researchers, biologists and chemists, were looking to design small molecule bromodomain 4 (BRD4) inhibitors. We needed reliable predictions to guide the design of these new compounds, particularly to improve the metabolic stability and solubility compared to existing structures.
How did StarDrop help you in this drug discovery project?
SZ: We used StarDrop extensively throughout this project. StarDrop’s ADME QSAR module provided essential solubility and metabolic stability predictions, alongside a number of other property models. For example, to help comply with FDA safety regulations, we were able to examine StarDrop’s hERG inhibition predictions for any potential cardiotoxicity.
To further enhance our toxicity predictions, we recently subscribed to the Derek Nexus module within StarDrop. This provides knowledge-based prediction of 40 key toxicity endpoints. The toxicity predictions are relevant to pre-clinical studies, so support us in prioritising the best candidates to progress within our projects.
We obviously corroborate these predictions with experimental tests. We have found that the Glowing Molecule™ feature, which highlights the regions of the molecule which may be contributing to toxicity (or other selected predicted properties), is consistent with our medicinal chemistry knowledge.
This simple, clear visualisation is very valuable in guiding our molecular design.
How else does StarDrop support you?
SZ: We use a number of features in addition to ADME QSAR. Firstly, if we have experimental data around a property for which there are no pre-built models, we can build our own custom models. And the R&D team at Optibrium have been very open and helpful as we’ve worked on these custom models. They particularly mentioned that even more powerful algorithms will be incorporated in future to support custom models, which I very much look forward to. We also use StarDrop’s SeeSAR modules for binding mode prediction. For projects where we know the protein targets, especially if we have co-crystal structures of parent compounds, we can use SeeSAR for targeted design of specific analogues.
Our design processes are further enhanced using StarDrop’s molecule enumeration features, such as R-group variation. Using both custom-designed and in-built StarDrop libraries, we can easily enumerate synthesisable compounds with known reactions and commercially available starting materials.
So, what’s so special about StarDrop?
SZ: Our drug discovery team is made up of both computational chemists, like me, and experimental scientists. Therefore, we need software that is quick and easy to use for experimental scientists. StarDrop fits the bill – it’s very user friendly and convenient.
It’s even more convenient now, having recently switched to the cloud version of StarDrop, so our team can access it everywhere, even on their phones. That flexible access is extremely valuable.
Of course, access is nothing without the software itself being useful and reliable. The guidance of the predictive models, particularly when used with the Glowing Molecule to display exactly which parts need to be targeted, is very helpful. It lets us pinpoint exactly which part of the structure needs to be modified to improve our property profile. This is really useful compared to other tools, which just provide numerical prediction values.
In addition to the software itself, there’s the professional, fast customer service.
Any issues we might have, the support team is able to provide satisfying solutions to very quickly. For simple questions, I can get an answer from the Application Scientists within 20 minutes. This, I really appreciate.
I also like that new features keep being released. The software is actively developed to meet our needs and the needs of the community.
How does StarDrop transform your discovery?
SZ: Without prediction tools such as StarDrop, each project would be very time-consuming and inefficient. Access to StarDrop also saved our time on developing in-house scripts for the same function. Without easy access to predictive models and SAR analysis, our medicinal chemists would likely have to synthesise vastly more molecules with less targeted designs, wasting time and money.
Where can we find out more about this anti-inflammation project?
SZ: The results of this particular project (with StarDrop method details available in the Supplementary Information) are published in Bioorg. Med. Chem. Lett.
Interested in trying StarDrop?
Get in touch with the team to discuss your specific projects and needs, and book a software demo.
More drug discovery resources
How can I make the most of my predictive models for drug discovery?
What’s the purpose of a predictive model? What’s the value of predictive models for drug discovery? Most of the undergraduate…
How do StarDrop and LiveDesign compare?
Finding the right fit There’s no one-size-fits-all solution when it comes to data visualisation and analysis tools for drug development…
