Running Single-Cell RNA Sequence Analysis on the dkNet Texera Cloud Platform

This guide walks you through how to perform single-cell RNA sequencing analysis on texera.dknet-ai.org. The platform provides two workflows you should run in order:

1. Step 1 – CloudBioMapper
2. Steps 2–7 – Single-Cell RNA-Seq Analysis

The two workflows represent the common single-cell RNA sequence analysis pipeline, covering both alignment and downstream processing.

Step 1: CloudBioMapper Workflow

The CloudBioMapper workflow performs read alignment using the STARsolo package. It requires three inputs:

1. FASTQ Files
   • These are your raw reads to align.
   • Use the sample dataset sample-fastq as a reference.
   • File names must follow this format:
     [SampleName]_S{sampleNumber}_L00{laneNumber}_{ReadType}_001.fastq.gz
2. Reference Genome
   Use the sample sample-reference dataset or upload your own.
3. Compute Cluster
   This is where the alignment runs.

Before running the workflow:

1. Go to “CloudBioMapper Clusters” in the left panel.
2. Click Create Cluster.
3. Set:
   • Name
   • Machine type
   • Number of machines (more machines = faster, but higher cost)
4. Creation takes ~1 minute. A green check mark confirms readiness.

Running the workflow:

1. Select your FASTQ dataset, reference genome, and cluster.
2. Click Run.
3. Execution takes ~5 minutes.
4. Once complete, results appear in the result panel with four output columns:
   • Sample
   • features.tsv.gz
   • barcodes.tsv.gz
   • matrix.mtx.gz

Exporting Results for the “Step 2–7 Workflow”

Before you proceed, export the alignment results:

1. Right-click the CloudBioMapper operator.
2. Select “Export Result.”
3. Set:
   • Export Type: Binary Format (.arrow)
   • Destination: Dataset
4. Tip: Create a new empty dataset if you want to keep the results separate.
5. Click Export.

Step 2–7: Running Downstream Analysis

1. Open the Steps 2–7 workflow.
2. Click the Arrow File Scan operator.
3. Select the dataset you exported from Step 1.
4. Save the configuration.
5. Click Run to begin processing.

This workflow will perform quality control, filtering, normalization, and additional analysis.

Quick Tips

• Cluster Sizing: Use more machines for large datasets to reduce runtime.
• File Naming: Ensure your FASTQ files match the required format.