Skip to main content
Figure 1 | BMC Proceedings

Figure 1

From: A novel low-parameter computational model to aid in-silico glycoengineering

Figure 1

(a)Glycosylation is the outcome of a complex reaction network where the glycoprotein is successively processed by several enzymes in a sequence of Golgi compartments.(b)In the model glycosylation is conceptualized as a random walk (Markov chain) starting in the initial Man9GlcNAc2 glycan and reaching a certain secreted glycan with a certain secretion probability. (c) Secreted glycans are modelled as absorbing states of the Markov chain that transition to themselves with probability 1, once they have been reached. To fit the unknown transition probabilities, the user inputs the experimentally derived glycoprofile of the protein of interest. (d) The deduced transition probabilities are arranged in a Markov transition matrix that gives the probabilities of transitioning from any glycan (rows) to any other (columns). (e)To model an enzyme knockout, the corresponding transition probability is set to 0. Optimization is used to adjust the alternative transition probabilities to maintain a probability sum of 1 for every glycan affected by the knockout. (f)Wildtype and Fut8 knockout glycoprofile of CHO-S secretome (black bars), together with the glycan frequencies as predicted by the model (red bars).

Back to article page