Dr James G Wright PhD, Scientist

I originally studied Mathematics at Cambridge University, and subsequently obtained a PhD entitled "Nonlinear Hierarchical Modelling of Chemotherapeutic agents". Very early in my career, I realized that relying entirely on statistics (the science of analyzing data) was insufficient. My earliest analyses were naive in that they pretended that all the information one could ever need was contained in the dataset, and ignored all of the existing biological knowledge. I quickly learnt that knowing a great deal about equations could not substitute for understanding what the equations imply in terms of biology and pharmacology.

Over the following decade, I have developed hundreds of pharmacokinetic and pharmacodynamic models whose structure is derived from biology and pharmacology, whilst retaining statistical validity. From the established biomarkers in cardiovascular medicine, to new innovations in neuroscience and oncology, there is always insight in the minds of scientists that can strengthen and define appropriate mathematical models. Models that make sense to scientists and clinicians, and whose parameters represent real biology, are far more useful than empirical fits to the data. In this way, the data analysis approximates reality rather than attempting to warp reality into a standardized statistical analysis.

Wright Dose Ltd acts as a translator between the language of scientists and the language of computers, and by bringing the two closer together, we can account for the complexity in the data and evaluate hypotheses that represent what scientists are actually thinking.

Dr David Cristinacce PhD, Scientist

I also studied Mathematics at Cambridge, and after working briefly as an actuary, returned to academia in the field of computer vision. My PhD was titled "The Automatic Detection of Facial Features in Grey Scale Images" and I collaborated with a major car manufacturer on technology that detected driver fatigue. In 2009, I left the University of Manchester and joined Wright Dose Ltd to work on models of complex biomarker systems, primarily in WinBUGs. My interests include automation of complex analyses via high throughput computing, time series models of multivariate categorical data and methods for evaluating the quality of model predictions.