stlearn is a Python package designed for the analysis of spatial transcriptomics data. It serves as a toolkit for researchers aiming to unravel complex biological processes within intact tissues by integrating various data types. The primary purpose of stlearn is to combine spatial distance, tissue morphology, and gene expression measurements, providing a more accurate model of underlying tissue biology. This integrative approach enhances understanding of gene expression in its precise cellular and anatomical context, moving beyond traditional methods that often lose crucial spatial information.
It enables investigations into cell-type identification, spatial cell trajectory reconstruction, and cell-cell interactions within undissociated tissue samples. By leveraging multiple data dimensions, stlearn helps bridge the gap between gene expression patterns and their physical manifestation in tissues. It helps decipher the intricate molecular landscapes that govern both healthy biological functions and disease states.
Understanding Spatial Transcriptomics
Spatial transcriptomics is a technology that allows scientists to measure gene activity while preserving the precise location of these activities within a tissue section. Unlike older methods that required dissociating tissues into individual cells, which resulted in the loss of spatial context, spatial transcriptomics maintains the architectural integrity of the sample.
The ability to retain spatial information is important for understanding biological processes and diseases. Cells do not function in isolation; their behavior is heavily influenced by their immediate neighbors and the broader tissue environment. For instance, in a tumor, the gene expression profile of a cancer cell can be influenced by surrounding immune cells or stromal cells, and understanding these interactions requires knowing their spatial proximity.
The data generated by spatial transcriptomics technologies, such as 10x Genomics Visium or NanoString GeoMx, are complex. They involve large amounts of gene expression data and high-resolution image data, creating a multi-dimensional dataset that is challenging to analyze manually. This complexity highlights the need for specialized computational tools like stlearn, which can effectively integrate and interpret these diverse data types to extract meaningful biological insights.
Core Capabilities of stlearn
stlearn performs analyses on spatial transcriptomics data by integrating various data modalities. One of its strengths lies in data integration, where it combines gene expression profiles with spatial location and tissue morphology information. This allows for a more holistic view of the tissue, moving beyond just gene lists to understanding how gene activity is organized in space.
The tool facilitates spatial domain identification, which means it can pinpoint distinct regions within a tissue based on their unique gene expression patterns and spatial organization. For example, stlearn can identify different layers in the brain or specific compartments within a tumor microenvironment. This is achieved by performing spatial clustering procedures that segregate cell types within a tissue.
stlearn also supports cell type mapping. It uses a two-step spatial clustering procedure, adapting methods commonly used in single-cell RNA sequencing analysis, to identify cell types and cellular phenotypes based on normalized gene expression values. Furthermore, stlearn enables the analysis of cell-cell interactions. It identifies “hotspots” in the tissue where high ligand-receptor interaction activity and diverse cell type co-localization occur, suggesting regions of significant cellular communication.
Applications and Research Impact
stlearn has applications across biological and medical research. In cancer biology, for example, stlearn is being used to understand the tumor microenvironment with spatial detail. It helps identify how different cell types, such as immune cells and cancer cells, interact within a tumor, which can reveal mechanisms of tumor progression or resistance to therapy. By mapping gene expression spatially, researchers can uncover the precise cellular neighborhoods that drive disease.
The tool also contributes to developmental biology by enabling the mapping of tissue development and organogenesis. Researchers can track changes in gene expression and cell organization over time and space, providing insights into the processes that guide the formation of complex structures from simple beginnings. This helps to understand how cell lineages differentiate and organize to form functional tissues and organs.
In neuroscience, stlearn is applied to study brain architecture and neuronal circuits. It allows for the examination of gene expression patterns within specific brain regions, helping to delineate different neuronal populations and their spatial relationships. This can shed light on the molecular underpinnings of neurological disorders and normal brain function. The tool is also used in immunology to investigate immune cell infiltration and organization within tissues, understanding how immune responses are coordinated in a spatial context. Insights from stlearn contribute to new discoveries, aiding improved diagnostics by identifying spatial biomarkers and informing targeted therapies for various diseases.