Manipulation of biological particles including pulling, and holding-and-indenting them, using the atomic force microscope (AFM) has attracted enormous interests. High deformability and vulnerability of biological particles, especially cells, make moving toward the target point inside complex biological environments with the least invasion the most critical factor. In this article, the optimal path of the particle movement is determined by considering the mechanical and morphological properties of the biological cell. Furthermore, the shortest path with the least amount of cell deformation is determined by using the equations of 3D manipulation of spherical viscoelastic particles and genetic algorithm (GA). Eventually, the final path is determined considering the mechanical properties of breast cancer cells by applying different constraints such as folding factor and the particle's roughness.Results reveal that increasing the number of constraints raise the needed time to find the optimal path. Additionally, the maximum time belongs to the spherical particle in the presence of folding. As a result, the total path planning times for the smooth, rough, and folded spherical particle are 59.386, 129.578, and 214.404s, respectively. Various optimal pathfinders are used, to reduce calculations and speed up the process, as well as obtaining the correct answer with high certainty. Comparing the error files founded for three methods including cellular learning Automata, Dijkstra, and GA, the third method has the best performance in the lowest error rate. Using the GA, the error rate can be reduced by 40%, compared to the cellular learning Automata method. Furthermore, comparing the cellular learning Automata method used in previous studies, it can be seen that not only the results are correct, but also less time spent, at the practically identical situation, on finding the optimal path for this algorithm.