Proteolysis-targeting chimera (PROTAC), as a paradigm of proximity-inducing compounds, uses a unique bifunctional small-molecule design strategy, namely, building the structure with three components—a warhead targeting the protein of interest (POI) or target protein, an anchor ligand binding to the E3 ubiquitin ligase (E3 ligand), and a linker connecting the two ligands [1, 2, 3]. To induce target protein degradation, the two ends of a PROTAC need to simultaneously bind to the POI and E3 ligase correspondingly, with the linker to adjust the spatial arrangement of the two proteins. Upon forming the POI-PROTAC-E3 ligase ternary complex, the E3 ligase catalyzes ubiquitination of the POI (Figure 1a), leaving the ubiquitination-tagged protein to be recognized by intracellular proteasomes and subsequently degraded [4]. This event-driven catalytic mechanism endows PROTACs with revolutionary advantages—unlike traditional inhibitors that function by continuously occupying the active site of target protein, PROTACs only need to facilitate the formation of the degradation complex transiently to trigger irreversible protein degradation [5, 6, 7]. This mechanism enables PROTACs to function on traditionally “undruggable” targets, for example, scaffold proteins lacking well-defined active pockets or mutated oncoproteins that exhibit decreased sensitivity to the traditional inhibitors [8,9], therefore significantly extending the target space.

Notably, the degradation efficiency of a PROTAC is intimately associated with the dynamic characteristics of the induced POI–PROTAC–E3 ligase complex [10, 11, 12, 13]. Numerous studies have demonstrated that structure stability of the PROTAC-mediated ternary complex serves as a key factor for evaluating its degradation efficiency [14,15] and that the effective establishment of protein–protein cooperativity constitutes a prerequisite for successfully assembling the degradation complex [16,17]. If the protein–protein interaction (PPI) within a PROTAC-mediated ternary system possesses cooperative advantage (positive cooperativity, α > 1), the binding of PROTAC to one protein (POI or E3 ligase) will enhance its affinity to the other (E3 ligase or POI), making the system prefer to form a POI–PROTAC–E3 ligase ternary complex [18]; conversely, when negative cooperativity happens (α < 1), a PROTAC is inclined to preferentially form nonfunctional binary complexes with the target protein or E3 ligase, leading to the significant reduction of degradation activity [18,19] (Figure 1b). Moreover, in addition to the stabilization and cooperativity effects, the residence time (characterized by 1/Koff) of the ternary complex also directly influences the ubiquitination efficiency of the POI, where a sufficiently long residence time guarantees the adequate ubiquitination labeling of the POI [20].

To characterize the key factors responsible for PROTAC’s degradation efficiency, various computational approaches have been developed in addition to experimental methodologies. However, the precise calculation of these key factors still faces several challenges: (1) the flexibility of PROTAC itself can induce differential PPI patterns within the ternary complex, leading to significant fluctuation in degradation efficiency [21,22]; (2) the formation, functioning, and degradation of PROTAC degradation machinery complex involve complicated dynamic cooperation between multiple components, with the key steps still not fully elucidated [23]; (3) moreover, the severe scarcity of publicly available crystal structures of PROTAC ternary complexes also seriously limits the development and application of artificial intelligence (AI)– or machine learning (ML)–based methodologies [24,25]. To address these challenges, this review, focusing on molecular dynamics (MD)–based technologies, systematically evaluates the cutting-edge methodologies reported in recent years. Through this ‘dynamic’ perspective, we aim to decipher the operational principles and design strategies governing PROTAC functionality.