Deep Visual Proteomics (DVP) is a groundbreaking approach for analyzing cellular phenotypes, developed using Biology Image Analysis Software (BIAS). It combines advanced microscopy, artificial intelligence, and ultra-sensitive mass spectrometry, considerably expanding the ability to conduct comprehensive proteomic analyses within the native spatial context of cells. The DVP method involves high-resolution imaging for single-cell phenotyping, artificial intelligence-driven cell segmentation, and automated laser microdissection to isolate cellular or subcellular regions of interest accurately.
A key feature of DVP is its integration of imaging and proteomic technologies. These allow for the identification of cell types and delineation of distinct states, based on AI-defined features. By preserving spatial information alongside molecular insights, DVP provides a powerful analytical tool. It has applications in research and clinical diagnostics related to cell and disease biology, including the study of single-cell heterogeneity and characterizing proteomic differences in disease tissues such as melanoma and salivary gland carcinoma.
The image processing and single-cell isolation workflow in DVP is comprehensive. It begins with high-resolution scanning microscopy, capturing whole-slide images that are processed in BIAS, which supports various microscopy formats and uses deep-learning algorithms for precise segmentation of cellular components like nuclei and cytoplasm.
DVP is especially effective at characterizing functional differences among distinct cells at the subcellular level. In one research example, scientists used deep learning-based segmentation to isolate and analyze individual cells and nuclei from an unperturbed cancer cell line. The resulting proteomic profiles were distinctive and highly reproducible across different cell classes, contributing insights into cell cycle regulation and potential cancer prognostic markers.
Applied to archived disease tissue samples, DVP revealed significant proteomic differences between normal and cancerous cells in a sample of salivary gland acinic cell carcinoma tissue. Similarly, in melanoma research, DVP identified distinct proteomic signatures linked to disease progression and prognosis.
DVP offers potential for extensive applications across diverse biological systems capable of microscopic imaging, such as cell cultures and pathology samples. It allows the rapid scanning of slides to isolate rare cell states and study the proteomic composition of the extracellular matrix. This technology has significant potential, particularly in oncology, where it amplifies digital pathology by providing a thorough proteomic context. As a result, DVP could facilitate more precise molecular subtyping of diseases, thus assisting in clinical decision-making.
