An optimized, high-throughput workflow for the collection, processing, and visualization of histology data in...

In an era where the diversity and quality of imaging modalities is rapidly increasing, it may seem counterintuitive to promote classical histology as a critical skill. It is a mistake, however, to assume that the heuristic potential of these high-resolution histological data is stagnant. Deep learning algorithms have emerged as an efficient tool for converting and quantifying the cellular resolution of 2D histology sections as detailed 3D models, capable of being integrated with a diversity of multi-omics data. Such analytical innovation requires large numbers of high-quality slides whose construction faces a variety of technical challenges. These challenges are exaggerated for developmental and evolutionary biologists, for whom ontogeny and phylogeny are critical variables that require additional sampling. Our goal is to provide a protocol optimized for the thin-section histology of vertebrate embryos, detailing best practices for sample collection, processing, and slide preparation. We hope that by: (1) synthesizing a scattered methodological literature that often excludes embryological tissues, and (2) recommending adjustments to common techniques like dehydration, xylene infiltration, and sectioning, other researchers may bypass the frustrating and time-consuming problems we encountered and move quickly to producing the high-quality histological data that modern developmental biology is likely to demand.