Abstract
Glycans, nucleic acids and proteins are three major classes of natural biopolymers. The extremely high diversity of isomerization makes structural elucidation of glycans the most challenging job among three classes. In the past few years, the single molecule sensing technique based on nanopores has achieved great success in sequencing of DNA. Inspired by this, it is potential to sequence glycans in the similar manner. Herein, SiN(x) nanopores were used to identify nine common monosaccharides in mammals. Each monosaccharide showed characteristic blockage current, which roughly increased with the increase of its molecular weight. In order to distinguish nine monosaccharides, several machine learning models were trained and tested, of which the highest F1 value was 1. These results illustrated that nine common monosaccharides in mammals could be clearly identified and discriminate by our method combining solid-state nanopores and machine learning. As far as we know, this is the first report that monosaccharides can be sensed and distinguished by solid-state nanopores. Our work showed the great potential of solid-state nanopores in glycan sequencing, and would lay the foundation for solid-state nanopore-based glycan sequencing.