This tutorial will use a pre-loaded dataset in scExplorer, the 3K PBMC data from 10x Genomics. To begin, go to the Upload tab in the side menu (1). If you want to explore one of the pre-existing models, click the Tutorial button (4), where you can choose among four datasets: PBMC, mouse cortex, zebrafish cranial neural crest, and a velocity-ready mouse pancreas dataset.

If you prefer to upload your own data, first define an Analysis Name (3), select the Species (6) — human, mouse, zebrafish, or others — and choose the Gene ID format (7), either Symbol or Ensembl ID. Optionally, provide an Email (8) to receive a notification once the upload is complete, along with the UUID of the analysis. Then drag and drop your files or select them using the clicking here link (9), and press the Upload button (10). If you have a previous run, you can resume it by entering its UUID in Load Run with UUID (11); the analysis will continue from where it left off.

In addition to uploading new data, scExplorer includes the option to reproduce past analyses. In the Reproduce Results section (12), you can set parameters manually with Set Parameters (1) or download a configuration template using Download Config Template (2). If you already have a configuration file, upload it with Select File (3). For Preprocessing, define thresholds such as Min Genes (4) to filter out cells that express very few genes, Min Cells (5) to discard genes detected in only a few cells, and the MT Threshold (7) to remove cells with high mitochondrial content. You can also enable or disable Doublet Detection (8) to automatically identify potential doublets.

For Embedding, set the Number of Highly Variable Genes (1) for dimensionality reduction, choose the HVG Method (2) such as Seurat or Scanpy, adjust the Number of Principal Components (3), define the Number of Neighbors (4), and control cluster granularity with Leiden Resolution (5).

For Differential Expression Analysis, specify the Number of Genes (1) to report per cluster or condition, select the Statistical Method (2), and run the workflow with Run Pipeline (3). Through this sequence, scExplorer lets you upload and analyze new datasets, revisit previous runs using UUIDs, or ensure reproducibility by setting or reusing parameters. This ensures consistent analyses and simplifies sharing across datasets and collaborators.

After loading the dataset, you will see a Dataset Summary (1) and three quality control (QC) plots. The left plot shows the number of genes per cell (2), the middle plot shows the total counts/UMIs per cell (3), and the right plot shows the percentage of mitochondrial genes per cell (4). Each dot represents a unique cell. Click Continue to Preprocessing (5) to proceed, and use Export (6) to download the QC plots.

You can customize the look of your plots using the Plot Theme Customization dialog. Start in the Presets tab (1), which offers quick themes such as Default, Light, Colorblind, High Contrast, Greys, and Vibrant (2). The Live Preview area (3) shows how scatter and violin plots will look. Click Apply Theme (4) to confirm or Reset (5) to restore defaults.

In the By Plot Type tab (1), adjust settings for each plot style (2). For Violin Plots you can change colors, opacity, and line width; for Scatter Plots you can set colors, marker size, and opacity. Previews update live as you make changes.

In the Global Settings tab (1), control overall styling (2) such as plot background, grid color and visibility, font family, title size, axis label size, and font color. These settings apply across all plots to keep a consistent look.