29 May 2024 | Linlin Yang, Qian Tang, Mingzhi Zhang, Yuan Tian, Xiaoxing Chen, Rui Xu, Qian Ma, Pei Guo, Chao Zhang, Da Han
This article presents a spatially localized DNA integrated circuits-based classifier (DNA IC-CLA) for cancer diagnosis. The DNA IC-CLA employs a two-dimensional DNA origami framework and localized processing modules to execute arithmetic operations (multiplication, addition, subtraction) for efficient linear classification of complex patterns of miRNA inputs. The system demonstrates accurate and faster (about 3 hours) cancer diagnosis in synthetic and clinical samples compared to traditional freely diffusible DNA circuits. The DNA IC-CLA's neuromorphic architecture enables precise concentration-based transformations and robust arithmetic operations, enhancing computational speed and accuracy. The study also includes an in silico-trained classifier model using miRNA-seq data from the TCGA database, achieving high sensitivity, specificity, and accuracy in diagnosing non-small cell lung cancer (NSCLC). The DNA IC-CLA's modular design and enhanced physiological stability make it a promising tool for in vivo applications and future diagnostic platforms.This article presents a spatially localized DNA integrated circuits-based classifier (DNA IC-CLA) for cancer diagnosis. The DNA IC-CLA employs a two-dimensional DNA origami framework and localized processing modules to execute arithmetic operations (multiplication, addition, subtraction) for efficient linear classification of complex patterns of miRNA inputs. The system demonstrates accurate and faster (about 3 hours) cancer diagnosis in synthetic and clinical samples compared to traditional freely diffusible DNA circuits. The DNA IC-CLA's neuromorphic architecture enables precise concentration-based transformations and robust arithmetic operations, enhancing computational speed and accuracy. The study also includes an in silico-trained classifier model using miRNA-seq data from the TCGA database, achieving high sensitivity, specificity, and accuracy in diagnosing non-small cell lung cancer (NSCLC). The DNA IC-CLA's modular design and enhanced physiological stability make it a promising tool for in vivo applications and future diagnostic platforms.