Séminaire des Doctorant·e·s :

Le 12 octobre 2022 à 17:30 - Salle 109

Présentée par Chen Junyi - IMAG

Coagulation cascade based on gradients and Ordinary Differential Equations

During the past few decades, the research about computational coagulation which is faster and more convenient than experimental methods has developed. We developed an optimizer for coagulation cascade to find the best parameters, based on gradients and Ordinary Differential Equations(ODE). Normally, the dynamics of thrombin, which will convert fibrin to fibrinogen, is the key specie in coagulation. Two models has been optimized with the purpose to achieve the ideal generation of thrombin. The simple model had four variables, while the complex one had 10. It is found that for complex models, there were many local optima and the time cost was a little expensive. Therefore, a faster method for calculating the gradients is applied as well and the complexity of the model is analyzed, which mainly focused on the local minima.