Established in 2000, DesertSci was founded by Dr Neil Taylor an experienced Molecular Modeller, to remedy what he saw as a lack of usable and effective software for medicinal chemistry. Much of the existing software didn’t fully meet the needs of computational chemists in medicinal chemistry project support; it was over-priced and overly complex to use, limiting it to expert users only; and technical support was poor.
DesertSci was established to solve these issues by web-enabling computational chemistry and chemical informatics. More specifically, we enable highly skilled scientists to internally share their data and methods with minimal effort. We are data focussed and knowledge based. That is, we integrate scientific expertise with data evidence to create systematic research software. By hiding sophisticated data and focussed computational methodologies behind streamlined web interfaces, we enable all scientists in multidisciplinary teams to access complex technologies, that were once the domain of expert modellers, at the click of a button.
Since its incorporation in 2000, DesertSci has grown slowly and solidly. We released the first production version of Proasis in 2001, following extensive collaboration with AstraZeneca. We then worked closely with other major clients such as GlaxoSmithKline, F.Hoffman-La Roche and Novartis AG on improving our software and interfaces to meet ever changing industry needs. In 2005 we embarked on a joint venture with F.Hoffman-La Roche to develop Scorpion, our ground-breaking method using cooperative networks to model tight ligand binding. This partnership enabled us to train our innovative methodologies on the highest quality data available. A paper, entitled ‘Rationalizing Tight Ligand Binding through Cooperative Interaction Networks’ detailing the methods involved in developing Scorpion was published in J Chem Inf Model. 2011 Dec 9. Scorpion, provides exciting insights into binding, particularly in difficult to understand molecular structures. DesertSci is currently working on Scorpion2, and anticipate that this will provide radical improvements in determining tight binding.
Other major accomplishments include the development of our MOS clustering methods (Spinifex) in 2004 and the release of BMOS (2016). BMOS uses a combination MCS/MOS approach to identify important structural activity relationships, in seemingly dissimilar molecules, and gain important insights that positively impact the design of synthetic candidates. This is all done on very large (>100k) datasets with reasonable job times and high levels of accuracy.
Proasis has seen constant evolution – with the popularity of high throughput screening, the need for a sophisticated system for organising and understanding the plethora of data has become increasingly relevant. Proasis now sits at the core of many of the world’s largest pharma companies’ infrastructure and is integral in taming this ‘data beast’. Smaller companies also understand its value and we have seen unprecedented uptake by these companies, in managing their exciting ‘hands on’ drug discovery. In 2017 Proasis4 was released and comprised a complete version overhaul, including an amazing new interface designed by our UX specialist, inbuilt graphics and extensive and powerful upgrades to our inbuilt computational methods.
Our small team of highly experienced and committed professionals comprises software developers, UX specialist, web experts and operations personnel who are all committed to working closely with our clients to meet their needs and develop innovative products to cater for the complexities of the modern world of drug discovery.
DesertSci continues to focus on methods development coupled with software usability. We have a strong pipeline of leading edge software products and many more new and inventive ideas to develop.
