The protein corona forms when proteins in blood, saliva or other biological fluids attach to the surface of nanoparticles, shaping how the body's cells recognize and respond to them. The phenomenon matters for drug delivery, diagnostics and any technology that introduces nanoparticles into a living system, since the protein layer that forms can determine whether the body treats the particle as harmless or flags it for attack.

Researchers led by Hongdong Hao analyzed two separate sets of publications: a general corpus of 1,212 papers spanning 2000 to 2025, and a specialized corpus of 310 papers focused on computational and AI-enabled approaches, running from 2011 to 2025. Hao said the team split the literature this way because the computational field, despite its growth, risked being obscured by the much larger volume of traditional experimental work if the two were analyzed together.

The analysis found that computational research has stayed closely tethered to the traditional field rather than branching off on its own. Papers using simulation and machine learning methods tend to appear in the same journals as experimental studies and frequently cite the same foundational work. Within that shared foundation, the two tracks have moved in different directions: the traditional literature has increasingly concentrated on nanomedicine and drug delivery applications, while the computational side has centered on machine learning and molecular dynamics techniques.

Hao said the paper was written to serve as an entry point into the field. "I hope the paper gives readers an accessible starting point for understanding where protein corona research has come from and where it may be going," Hao said.

Hao added that computational modeling alone will not be enough to make the field predictive. "Reliable prediction will still depend on good experimental data, consistent reporting, external validation, and close cooperation between computational and laboratory researchers," Hao said.