Tribus: Semi-automated discovery of cell identities and phenotypes from multiplexed imaging and proteomic data

Multiplexed imaging at single-cell resolution is becoming widely used to decipher the role of the cellular microenvironment in cancer and other complex diseases. To identify spatial patterns of single cells on a tissue, accurate cell-type phenotyping is a crucial step. This step is challenging due to (i) fluorescence noise and batch effects, (ii) segmentation artifacts, (iii) laborious annotation of ground truth, and (iv) difficulty in reproducing human-biased thresholding. Here we present Tribus, an interactive, knowledge-based classifier that avoids hard-set thresholds and manual labeling, is robust to noise, and takes fewer iterations from the user than current methods of labeling. Tribus has built-in visualization functions to gain insight into the input data and to evaluate the results. The Napari plug-in provides a user-friendly way to visualize the results and perform quality control. In this study, we validate the accuracy of Tribus by comparing its results to labels in public benchmarking datasets where manual cell type annotations are supported by the pathology community. We applied Tribus on a cyclic immunofluorescence (CyCIF) dataset, consisting of five matched ovarian cancer samples collected before and after neoadjuvant chemotherapy. Accurate cell-type phenotyping enabled a high-resolution analysis of cellular phenotypes, their spatial patterns, and their temporal dynamics during platinum-taxane chemotherapy. Tribus is provided as an easily embeddable open-source package and enables accurate phenotyping of single cells to facilitate biological discovery from highly multiplexed images. ### Competing Interest Statement The authors have declared no competing interest.