simulation

Long-Read Simulation Pipeline

This pipeline prepares reference data and performs long-read simulation using SQANTI-SIM. This directory contains the scripts required to simulate Long-Read transcriptomic data with controlled Differential Gene Expression (DGE), Differential Transcript Expression (DTE), and Differential Transcript Usage (DTU). It is organized into two sequential steps.

Pipeline Steps

Step 1: Reference Preparation

Script: 1.run_ref_prep.sh

This steps performs the following:

1. Alignment & Quantification: Aligns raw GTEx long-read data to the reference transcriptom using minimap2 and quantifies abundance using salmon.
2. Baseline Estimation: Runs gtex_info.R to generate baselineAbundance.rds and a list of expressed transcripts (txid.txt).
3. Subsetting: specific GTF, FASTA, and BED files for the identified transcripts.

Usage:

sbatch 1.run_ref_prep.sh
# Ensure this completes successfully before running Step 2

Step 2: Simulation

Script: 2.run_simulation.sh

This step performs the actual simulation:

1. Design: Runs sqanti-sim.py design to create the simulation index.
2. Differential Expression: Calls get_diff.R to establish DGE/DTU/DTE ground truth lists.
3. Dispersion: Calls simulate_dispersion.R to simulate biological variation across replicates.
4. Run Simulation: Executes sqanti-sim.py sim to generate synthetic FASTQ reads for Control and DE conditions.

Usage:

sbatch 2.run_simulation.sh

Dependencies

• Input Data: Expects raw data at ../../dataset/GTeX/long-read/sequence_data/
• Reference Files:
  • Transcript Reference FASTA: gencode.v44.transcripts.fa (Required for Step 1)
  • Genome Reference FASTA: GRCh38.primary_assembly.genome.fa (Required for Step 2)
  • Annotation GTF: gencode.v44.annotation.gtf
• Software & Versions:
  • minimap2 (2.26-r1175)
  • salmon (1.10.2)
  • samtools (1.19.2)
  • seqkit (2.5.1)
  • SQANTI3 (5.1.2)
  • SQANTI-SIM (0.2.1)