Investigation of thermodynamic properties of lattice and off-lattice models for proteins and polymers

有限长分子链的热力学行为的理解是非常重要的，因为对于杂聚物的构象状态的序列依赖类的识别是相当困难的。我们最近开发的多正则链生长方法能够在几乎所有温度下对晶格杂聚物进行采样。通过将其应用于蛋白质的简单晶格模型，序列和构象之间的关系将被分析，以及能量和结构量的波动是单体构象重排的指标。这将被转移到非格模型，其中应用了多规范采样和适当适应的链增长算法。对于这些模型，我们集中我们的注意力在低温下的玻璃样行为的研究。为了对其进行充分的描述，必须引入和研究合适的系统或“序”参数，如结构重叠参数。此外，我们的方法的应用程序有限长度的晶格均聚物与（伪）一阶的d größer als 3维的过渡是有趣的，因为该算法规避构象障碍与不太有利的状态。

The understanding of the thermodynamic behaviour of molecular chains with finite length is of extraordinary importance, since for heteropolymers the identification of the sequence-dependent classes of conformational states is quite difficult. Our recently developed multicanonical chain-growth method is capable to sample lattice heteropolymers at almost all temperatures. By applying it to simple lattice models for proteins, the relationship between sequence and conformation shall be analysed as well as fluctuations of energetic and structural quantities being indicators of conformational rearrangements of the monomers. This is to be carried over to off-lattice models, for which multicanonical sampling and a suitably adapted chaingrowth algorithm are applied. For these models we focus our attention to the study of the glass-like behaviour at low temperatures. For its adequate description, suitable system or "order" parameters, such as the structural overlap parameter, have to be introduced and studied. Furthermore, the application of our method to finite-length lattice homopolymers with (pseudo-)first-order-like transitions in d größer als 3 dimensions is intriguing, since the algorithm circumvents conformational barriers associated with less-favoured states.