bio-spatial-transcriptomics-spatial-preprocessing

Preprocess spatial transcriptomics data by performing quality control, filtering, normalization, and feature selection to output a PCA-ready .h5ad file. This pipeline calculates spot QC metrics, filters genes, and identifies highly variable genes using Scanpy and Squidpy.

Identifying spatially variable genes in Visium data involves calculating QC metrics and applying combined feature selection. This Skill uses Squidpy and Scanpy on AnnData objects to perform SVG identification alongside HVG selection for robust downstream analysis.

Yes, this preprocessing pipeline works directly with AnnData objects in Python. It requires raw counts to be stored in adata.layers['counts'] and uses Squidpy with Scanpy to execute QC, normalization, and feature selection before outputting a preprocessed .h5ad file.

Calculating QC metrics for spatial omics spots involves using Scanpy and Squidpy to evaluate spot quality and mitochondrial percentage. The pipeline filters spots and genes based on these metrics to ensure robust data quality before normalization.

Yes, you can use Squidpy and Scanpy for PCA-ready spatial data normalization. The workflow normalizes filtered Visium-like data, performs HVG and SVG identification, and prepares the AnnData object for principal component analysis.

Spatial transcriptomics preprocessing requires raw counts in adata.layers['counts'] to preserve original data integrity. Maintaining this raw layer allows the pipeline to accurately perform normalization and feature selection to yield a preprocessed .h5ad file.