Corn stover is the most abundant form of crop residue that can serve as a source of lignocellulosic biomass in biorefinery approaches, for instance for the production of bioethanol. In such biorefinery processes, the constituent polysaccharide biopolymers are typically broken down into simple monomeric sugars by enzymatic saccharification, for further downstream fermentation into bioethanol. However, the recalcitrance of this material to enzymatic saccharification invokes the need for innovative pre-treatment methods to increase sugar conversion yield. Here, we focus on experimental glucose conversion time-courses for corn stover lignocellulose that has been pre-treated with different acid-catalysed processes and intensities. We identify the key parameters that determine enzymatic saccharification dynamics by performing a Sobol's sensitivity analysis on the comparison between the simulation results from our complex stochastic biophysical model, and the experimental data that we accurately reproduce. We find that the parameters relating to cellulose crystallinity and those associated with the cellobiohydrolase activity are predominantly driving the enzymatic saccharification dynamics. We confirm our computational results using mathematical calculations for a purely cellulosic substrate. On the one hand, having identified that only five parameters drastically influence the saccharification dynamics allows us to reduce the dimensionality of the parameter space (from nineteen to five parameters), which we expect will significantly speed up our fitting algorithm for comparison of experimental and simulated saccharification time-courses. On the other hand, these parameters directly highlight key targets for experimental endeavours in the optimisation of pre-treatment and saccharification conditions. Finally, this systematic and two-fold theoretical study, based on both mathematical and computational approaches, provides experimentalists with key insights that will support them in rationalising their complex experimental results.

玉米秸秆是最丰富的农作物残渣形式，可作为生物精炼方法中木质纤维素生物质的来源，例如用于生产生物乙醇。在这种生物精炼过程中，多糖生物聚合物的成分通常通过酶糖化分解成简单的单体糖，用于进一步下游发酵成生物乙醇。然而，这种材料对酶糖化的抵抗力导致需要创新的预处理方法来提高糖转化率。在这里，我们重点关注经过不同酸催化过程和强度预处理的玉米秸秆木质纤维素的实验葡萄糖转化时间过程。我们通过对复杂随机生物物理模型的模拟结果与我们准确再现的实验数据之间的比较进行索博尔敏感性分析，确定了决定酶糖化动力学的关键参数。我们发现与纤维素结晶度相关的参数和与纤维二糖水解酶活性相关的参数主要驱动酶糖化动力学。我们使用纯纤维素基材的数学计算来确认我们的计算结果。一方面，确定只有五个参数会极大地影响糖化动力学，使我们能够减少参数空间的维数（从十九个参数减少到五个参数），我们预计这将显着加快我们用于比较实验和模拟的拟合算法糖化时间过程。另一方面，这些参数直接突出了预处理和糖化条件优化实验努力的关键目标。最后，这项基于数学和计算方法的系统性和双重理论研究为实验学家提供了关键见解，支持他们合理化复杂的实验结果。