Human Absorbable MicroRNA Prediction based on an Ensemble Manifold Ranking Model

Area of Science:

• Molecular Biology
• Genetics
• Bioinformatics

Background:

• MicroRNAs (miRNAs) are short non-coding RNAs regulating human genes and biological processes.
• Previously thought to be solely endogenous, recent findings show humans can absorb exogenous miRNAs from diet.
• This raises questions about which dietary miRNAs enter circulation and exert functions.

Purpose of the Study:

• To develop a computational model for identifying human-absorbable exogenous miRNAs from dietary sources.
• To analyze dietary miRNAs and their features to predict circulation potential.
• To investigate the integration of exogenous miRNAs into human biological systems.

Main Methods:

• Developed an ensemble manifold ranking model.
• Analyzed 4,910 dietary microRNAs from 14 species using 1,120 sequence and structure-derived features.
• Selected 70 discriminative features to characterize circulating miRNAs and predict absorption.

Main Results:

• Identified 461 dietary microRNAs as transportable into human circulation.
• Validated model predictions through a milk-feeding study.
• Predicted and validated 26 viral miRNAs as transportable in external experiments.

Conclusions:

• The data-driven computational model effectively identifies transportable exogenous miRNAs.
• This approach offers a promising method to study circulating miRNAs without complex mechanistic investigations.
• Findings suggest dietary components can influence human molecular biology via exogenous miRNAs.