Molecular dynamics simulation [1] is a novel research methodology that has emerged alongside the progressive advancements in computer technology. Presently, this technique has evolved into a crucial approach within scientific investigations. Molecular dynamics represents an innovative method that integrates various fields including mathematics, physics, chemistry, and biology. Throughout the years, this approach has gained widespread utilization in molecular simulations. As a research technique in computational science, the simulation approach offers distinct benefits by offering detailed microstructural information with high spatial resolution and capturing structural dynamics with high temporal resolution. This makes it an essential tool for investigating the dynamic behavior of biomacromolecules at a microscopic level [[2], [3], [4], [5]]. In recent years, the field of computer science has witnessed significant advancements, leading to a remarkable increase in computational power. As a result, molecular dynamics simulations have emerged as powerful tools with diverse applications. Particularly in the domain of microscopic biology research, these simulations are gradually unveiling their potential by demonstrating high speed and strong correlation with experimental data [[6], [7], [8], [9]]. Molecular dynamics simulation methods possess the capability to monitor the evolution of system configurations throughout the simulation process. Presently, these techniques have successfully replicated systems at a timescale of one thousandth of a second [[10], [11], [12]].

Despite experimental methods being used to determine the mechanical properties of collagen molecules, the accuracy of these measurements is heavily reliant on sample quality and microscopic phenomena cannot be observed during experimentation. As a solution to these limitations, many scholars have turned to molecular dynamics simulation as a means of studying collagen molecules. In 2005, Lorenzo et al. [13] utilized the GROMACS software and the Single Molecular Dynamics (SMD) method to conduct stretching simulations on collagen-like molecules in a single-axis manner. The simulations were carried out at standard temperature and pressure conditions, employing the GROMOS96 Force Field. According to the simulated data, the Young's modulus of collagen molecules is approximately 4.87 GPa with a deviation of around 1.0 GPa, aligning well with existing experimental findings. In 2006, Buehler et al. [14] utilized the NAMD molecular simulation software to conduct simulation experiments on collagen molecules, investigating their mechanical properties such as stretching, compression, shearing, and bending. The results revealed that the mechanical characteristics of collagen varied depending on the rate at which it was stretched. Specifically, when subjected to stretching rates of 0.0001 A/step, 0.0002 A/step, and 0.001 A/step respectively, the Young's modulus of collagen was determined to be 6.99GPa, 8.71GPa, and 18.82GPa correspondingly. These findings suggest that the properties of collagen are influenced by the rate at which it is loaded. In 2009, a thorough examination was conducted by Matthew D. Shoulders and Ronald T. Raines [15] on three key elements pertaining to collagen fibers: the triple helix arrangement of collagen fibers, the mechanical characteristics exhibited by collagen fibers, as well as biomaterials derived from collagen. In 2010, Alfonso et al. [16] employed the GROMACS molecular simulation software to develop a coarse-grained representation of collagen molecules and elucidated their structural and mechanical characteristics. They determined the Young's modulus of a collagen molecule measuring approximately 8 nm in length, and their simulated findings were consistent with simulations and experimental results reported by other researchers. In 2014, a study conducted by Baptiste Depalle et al. [17] utilized molecular dynamics (MD) to investigate the impact of collagen crosslinking structure, density, and mechanical properties on the deformation of collagen fibers at a mesoscale level. The findings suggested that the stiffness of maximum deformation is influenced by both the type and density of collagen crosslinks when considering their mechanical properties as determining factors for collagen fiber stress-strain behavior. In 2015, Andrzej Mlyniec and colleagues [18] developed a simplified representation of collagen protein in three dimensions. They conducted simulations on the mechanical characteristics of collagen molecules in hydrated and dehydrated conditions using the LAMMPS molecular simulation software along with the Amber force field. In an aqueous environment, the Young's modulus of collagen remains approximately 7.4 GPa when the strain of collagen molecules is below 0.3, and it continues to be around 7.4 GPa even when the strain exceeds 0.3. However, in a non-aqueous environment, the Young's modulus increases from approximately 11.5 GPa for strains below 0.3 to about 64.8 GPa for strains exceeding this threshold. In 2016, Jee et al. [19] conducted a study utilizing molecular dynamics (MD) to examine the penetration process of water and ethanol molecules into dentin collagen fibers. Their hypothesis suggested that the removal of water molecules surrounding collagen by ethanol molecules is unlikely, and due to the narrow gaps between fiber layers, collagen penetration by ethanol molecules is limited. The simulation study confirmed that indeed ethanol molecules are unable to eliminate the layer of water molecules encompassing collagen; however, they can permeate through the interstitial regions between fiber layers but not through overlapping regions.

On a microscopic scale, the utilization of molecular dynamics simulation techniques has proven to be highly effective in examining the structure and dynamic information of biological macromolecular systems. Presently, due to limitations on a macroscopic level, it is challenging to observe the dynamic motion behavior of biomacromolecules during biological experiments at the microscopic level. Consequently, employing molecular dynamics simulation methods can enhance our analysis of microbehavior exhibited by biomacromolecules. This approach further aids in comprehending experimental phenomena that remain unexplained through traditional biological experimentation methods, enhances the overall process of conducting biological experiments, and facilitates predictions regarding experimental outcomes.

Collagen, a prevalent protein in vertebrates, makes up around 30% of all animal proteins [20]. It plays a crucial role as an extracellular matrix protein by imparting durability and structure to the skin, tendons, and bones of mammals. Moreover, it forms the primary constituent of these tissues.

The attention of scholars worldwide has been drawn to the remarkable supplementary healing effect of laser welding technology on skin wounds. Nevertheless, the precise mechanism behind collagen denaturation and wound healing caused by laser energy in incisions or wounds remains unclear. Type I collagen constitutes up to 80% of the composition in human skin. Based on preliminary experiments, we have observed a significant positive correlation between laser energy and the average temperature within the wound area during laser welding of skin incisions. By employing GROMACS to simulate the structure and force field of type I collagen molecules at different temperatures, we investigated the potential mechanism underlying collagen reconstruction and instantaneous tensile strength enhancement in laser-welded skin incisions through its variation pattern. However, the mere simulation of structural changes in type I collagen during the laser welding process is insufficient to fully elucidate the underlying mechanism of collagen reconstruction in laser-welded skin wounds. Therefore, this study employs finite element modeling using COMSOL software to simulate the temperature and stress fields generated during dual-beam laser welding, thereby providing a comprehensive understanding of the macroscopic mechanisms responsible for the remarkable healing effects observed in dual-beam laser-welded skin wounds. Prior to this, scholars have already conducted simulations of laser welding on skin. Ryabkin et al. [38] quantitatively evaluated the formation of weld seams and the area of tissue temperature necrosis during one beam laser welding on human skin using a CNTA computational model. They also assessed the composition of the used solder material, including bovine serum albumin (BSA), indocyanine green (ICG), and carbon nanotubes (CNTs), as well as the incident angle and pulse duration of the laser.

In summary, this study elucidates the structural and mechanical changes of type I collagen in skin tissue across different wound areas formed under dual-beam laser ablation energy conditions through experiments and molecular dynamics simulations. Additionally, with the aid of COMSOL's Multiphysics finite element simulation, a three-dimensional temperature field and stress-strain field during the dual-beam laser welding process for skin wounds are presented, providing insights into the healing mechanism at both microscopic and macroscopic levels.