<p>Protein conformational transitions are fundamental to the functions of many proteins, and computational methods are valuable for elucidating the transitions that are not readily accessible by experimental techniques. Here we developed accelerated sampling methods to calculate optimized all-atom protein conformational transition paths. Adaptively biased path optimization (ABPO) is a computational simulation method to optimize the conformational transition path between two states. We first examined the transition paths of three systems with relatively simple transitions. The ways to define reduced variables were explored and transition paths were built at convergence of the optimizations. We constructed the all-atom conformational transition path between the active and the inactive states of the Src kinase domain. The C helix rotation was identified as the main free energy barrier in the all‑atom system, and the intermediate conformations and key interactions along the transition path were analyzed. This is the first demonstration of the robustness of a computational method for calculating protein conformational transitions without restraints to a specified path. We also evaluated protein‑peptide interactions using both molecular dynamics simulations and peptide docking. Long unbiased simulations were used to evaluate Src‑SSP interactions and complex stability in both implicit and explicit solvent. Molecular docking was used to build possible protein‑peptide interaction models, using both Src regulatory domain SH2 and the kinase domain. Possible Src‑SSP complexes were built as the first Src‑substrate complex structure models.</p>