An AI analytics workbench for protein structural characterisation
Trajan Scientific and Medical is developing software for automation of Hydrogen Deuterium Mass Spectrometry, which can connect structure, behaviour and function of proteins, for understanding diseases and developing drug and vaccine treatments.
Modern AI techniques can provide powerful models for classifying and understanding protein structures, but expert supervision is required in the development, training and deployment of these models into automation scenarios.


Phase 1
Based on the GC-MS and LC-MS data output, Trajan Scientific and Medical developed a specialised software to analyse the signal, automatically detect chromatographic peaks, and calculate their area under the curve. Building on it, we aim to evaluate various and heterogeneous mathematical, statistical, and machine learning approaches designed to minimise error rates in peak detection. These methods will be assessed for their ability to improve accuracy, robustness, and computational efficiency.
