Industrial opportunities

UPS-CONA

Customisable real-time assay for discovery and quantification of activities in the ubiquitin system.

Drug combinations

Finding drug combinations which produce therapeutic benefit is a difficult task, usually resigned to chance or exhaustive network analysis and simulation by systems biologists. The CTB-Drug Combinations cheminfomatics platform enables the discovery of highly therapeutic synergistic drug combinations.

ASYN-CONA

Neurodegenerative diseases, such as Parkinson ́s or Alzheimer’s, are one of the greatest problems in modern medicine. In these diseases the misfolding and aggregation of proteins normally present in the body generates large fibrillary deposits that can be found in the brain of patients. The formation of prefibrillar small oligomers, considered the toxic species, has proven difficult to follow experimentally and harder still to inhibit. This novel bead-based imaging assay presents an efficient, effective and quantitative method of following their early formation, and enables a high-throughput screen to be performed for compounds able to restrict, or prevent the production of these toxic species, and presenting a potential route for controlling disease progression.

Compound archive annotation

We have developed a method to annotate a compound archive with predicted protein binders. This has the potential to drastically cut both time and costs associated with high throughput assay development, screening and consumables, prioritising only compounds predicted as likely to hit certain targets.

Bead Ring Evaluation and Analysis of Data Software (BREAD)

Bead-based confocal imaging assays have been demonstrated to be highly effective in the screening of compounds for inhibiting binding, enzymatic activity and aggregation. BREAD enables the automation of the analysis of these assays in a plate-based format, allowing rigorous statistical analysis of the image data.

CTB-Morph

The Auer group have developed a semi-automated, integrated and iterative process, “Morph”, to generate optimised peptidomimetics by inserting non-coding amino acids into peptides.

PuLSE-Phage Library Sequence Evaluation

We have developed software for the quality control of phage display libraries.

LogP prediction

The fields of drug discovery and medicinal chemistry have developed many rule sets in common use designed to direct efforts into the most promising chemical matter. The ever popular and famous Lipinski's rule of 5 along with others often include a logP term. Accurate prediction of this molecular property has long been the focus of many scientist, with vastly different approaches taken to make ever more accurate predictions. We present DRLogP (Dual Representation LogP), a logP prediction method designed to be most accurate within a range most relevant to medicinal chemistry efforts; that is, logP between 0 and 5 and molecular weight less than 800. DRLogP captures multiple representations of a molecule; local atom environments, the presence of described moieties along with the overall shape and electrostatics. With the molecule described at multiple levels, these representations are used by a neural network to predict to logP. DRLogP outperforms all tested methods, with a root mean squared error of prediction of 0.55.

Binding curve simulation

Accurate simulation of protein-ligand systems at equilibrium is at the heart of experimental planning and affinity determination. PyBindingCurve is a Python module capable of simulating and solving a wide range of multi-protein multi-ligand systems.