LigninFit: Stochastic Simulation Software for the Prediction and Analysis of Experimentally Observed Lignin Molecules

Lignin's complex and highly branched architecture plays a critical role in biomass utilisation. Yet, unravelling the intricate relation between lignin's measurable structural features and the underlying biosynthesis dynamics remains a major challenge. This difficulty stems from the variability in monolignol and bond type compositions, experimental datasets with typically distinct degrees of completeness, and extraction methods that differently impact molecular features. Together, this highlights the need for advanced numerical frameworks able to rationalise experimental data. We introduce LigninFit, a simulation software that integrates Lignin-KMC within an efficient and modular parameter optimisation procedure to infer lignin biosynthesis dynamics from bond distributions. Using experimental data from a curated set of 20 biomass samples, we reproduce their bond distributions following 3 versions of the fitting procedure, differing in the number of model parameters optimised. We then generate extensive in silico libraries of lignin structures for the best fit of each biomass and analysein depth, at both the ensemble and molecular levels, the properties of the resulting molecules in terms of bond distributions and branching degrees. Overall, LigninFit not only allows for predicting the complete bond distribution of incomplete datasets, and highlighting unexpected similarities and differences across biomass types, but also reveals which parameters most strongly impact the biosynthesis dynamics and which metrics are most suitable for discriminating among biomass types. Eventually, as a modular and open-source software, LigninFit paves the way for additional modelling developments, thereby contributing to guide experimental endeavours.