Tumorigenesis, metastasis, and the recurrence of cancer, which may result from the abnormal presence or activation of cancer stem cells (CSCs), are involved in disorders of exchanged matter (biomarkers), energy and information in living organisms. Rapid and sensitive detection and imaging of CSC biomarkers (such as CD133) are helpful for early diagnosis and therapeutic evaluation of tumors. Recently, a preliminary exploration of a few affinity molecules (like peptide-based probes) has just begun for chemical measurements and imaging of CSC biomarker CD133. However, a comprehensive analysis of the matter, energy and information in an artificial molecular system has not been demonstrated and applied to biosensing and disease diagnosis. In this study, a graphene-peptide-based fluorescent sensing system was constructed by utilizing a graphene oxide platform and a CD133-specific recognition peptide and comprehensively analysed with respect to matter (molecular events), energy (fluorescence) and information flow. The molecular event interaction networks in this system were further used to perform molecular logic computing, for the sensitive detection of CSC marker CD133 (with a linear range from 0 to 630 nM and a detection limit of 7.91 nM), and for an application involving targeting the imaging of cells and tumor tissues that highly express CD133 (with a detection limit of 1.1 × 103 cells per mL for CT26 CSCs). The present report will provide more opportunities for the development and design of molecular-level intelligent complex systems and will probably promote the development of artificial intelligent sensing and treatment systems, a molecular-level "Internet of Things", and artificial life.