Development of MIF/IL-1β biosensors for discovery of critical quality attributes and potential allergic rhinitis targets from clinical real-world data by intelligent algorithm coupled with in vitro and vivo mechanism validation.

There are still huge challenges from clinical real-world data to accurate targets and critical quality attributes (CQAs) for effective treatment of allergic rhinitis (AR). Here, we present a novel integrated strategy that biosensors and intelligent algorithms were used to angle AR targets and CQAs from clinical real world. Firstly, bagging and boosting partial least squares discrimination analysis (PLS-DA) and Monte-Carlo sampling were proposed to screen accurate AR targets. Macrophage migration inhibitory factor (MIF) and Interleukin-1beta (IL-1β) potential targets were obtained based on large-scale analysis of one thousand proteins and in-depth precise screening of seventy proteins. Furthermore, high electron mobility transistor (HEMT) biosensors were fabricated and successfully modified by MIF and IL-1β potential targets with a low detection concentration as 1 pM and quantitative range from 1 pM to 10 nM. Surprisingly, through MIF/IL-1β biosensors, we angled 5-O-methylvisammioside, amygdalin, and cimicifugoside three CQAs. The strong interaction was discovered among three CQAs and MIF/IL-1β biosensors with almost all KD up to 10-11 M. Finally, interaction among three CQAs and MIF/IL-1β biosensors were evaluated by in vitro and vivo experiments. In this paper, two critical potential targets and three effective CQAs for AR treatment were discovered and validated by biosensor and advanced algorithms. It provides a superior integrated idea for angling critical targets and CQAs from clinical real-world data by biosensors and informatics.