Cells at disease onset are often associated with subtle changes in the expression level of a single or few molecular components, making traditionally used biomarker-driven clinical diagnosis a challenging task. We demonstrate here the design of a DNA nanosensor array with multichannel output that identifies the normal or pathological state of a cell based on the alteration of its global proteomic signature. Fluorophore encoded ssDNA strands were coupled via supramolecular interaction with a surface functionalized gold nanoparticle quencher to generate this integrated sensor array. In this design, ssDNA sequences exhibit dual roles, where it provides differential affinities with the receptor AuNP as well as act as transducer elements. The unique interaction mode of the analyte molecules disrupts the noncovalent supramolecular complexation, generating simultaneous multi-channel fluorescence output to enable signature-based analyte identification via Linear Discriminant Analysis based machine learning algorithm. Different cell types, particularly normal and cancerous cells, were effectively distinguished using their fluorescent fingerprints. Additionally, this DNA sensor array displayed excellent sensitivity to identify cellular alterations associated with chemical modulation of catabolic processes. Importantly, pharmacological effectors, which could modulate autophagic flux, have been effectively distinguished by generating responses from their global protein signatures. Taken together, these studies demonstrate that our multi-channel DNA nanosensor is well suited for rapid identification of subtle changes in a complex mixture and thus can be readily expanded for point-of-care clinical diagnosis, high-throughput drug screening or predicting the therapeutic outcome from a limited sample volume.