Unsupervised Clustering of DNA Transmission Footprints Using MoS₂/WSe₂ Heterojunction

Quantum transport-based DNA sequencing is emerging as a promising technique in genetic analysis, offering fast, precise, and scalable decoding of genetic information, holding significant potential for applications in human biology and personalized medicine. Given the recent developments in supervised machine learning-coupled nanopore and nanochannel technology, predicting and classifying the labeled DNA nucleotides is now feasible with precision and accuracy. However, the next challenge arises as conventional analysis methods struggle to handle the vast amount of data generated by high-throughput DNA sequencing, particularly when dealing with complex spatial patterns in quantum transport readouts. Here, we propose an unsupervised machine learning approach utilizing a MoS₂/WSe₂ heterojunction to cluster the transmission footprints of unlabeled DNA nucleotides. Our approach streamlines the clustering of nucleotide signals, minimizing manual efforts while improving the speed and accuracy of nucleotide identification, making nanopore/nanochannel-based sequencing more scalable and precise. This study paves a new path toward clustering of DNA transmission readouts, providing a quick platform for the interpretation of genetic code.