A Guide to Whole Transcriptome CosMx® SMI Analysis

The WTX guide

Author
Affiliations

Evelyn Metzger

Bruker Spatial Biology

Github: eveilyeverafter

Preface

Someone once described the analysis of spatial transcriptomics as “rocket science,” and I remember balking at the comparison. In some ways, I thought, it’s far more complex. Humanity’s first steps on the moon were guided by software engineering that was itself a new frontier, yet it was built upon a foundation of physics understood for centuries. Even the first draft of the human genome, a monumental achievement from nearly a generation ago1, was a challenge of reading a single, linear code. Our task today is to interpret an entire library of those codes being read simultaneously across a bustling city of cells. This is the distinctly 21st-century frontier we now face.

This guide is the resource I wish I had when I first faced that frontier. It was born from my own journey of piecing together workflows from open-sourced code and gradually building an intuition for what this beautiful and complex data was revealing. My goal is not simply to provide a set of computational recipes, but to walk alongside you through a complete, end-to-end analysis of CosMx SMI Whole Transcriptome data. We will start with the raw files and move step-by-step through quality control, cell typing, and downstream analysis.

Along the way, we will move beyond static figures. This guide is designed to be an active learning experience, encouraging you to engage directly with the dataset to develop your own tactile feel for the analysis. This work is also dynamic in the approaches and content that is presented. New tertiary analyses are already in the works including ligand-receptor and trajectory analyeses. My hope is that by the end, you will have not only a robust workflow but also the confidence to adapt it to answer your own unique biological questions. That confidence is essential, because this is the landscape that Whole Transcriptome CosMx SMI data invites us into: a territory of discovery where the rules are still being written. Our challenge is not one of engineering known forces, but of interpreting a biological language we are only beginning to understand. This guide provides the tools to begin that translation.