Single-cell Long-read Simulator (SLSim)

Dependency

• pandas
• numpy
• Biopython
• tqdm
• Badread (error_simulator.py)

Introduction

1. SLSsim randomly sample cDNA from transcript reference and attach 10x cell barcodes, UMI, poly(dT) and Template switching oligo (TSO) to form perfect reads (i.e. reads with no error.): bin/perfect_read_generator.py
2. SLSsim using the error model and qscore modle from Badread (one of the depencencies) to introduce error into the reads (bin/error_simulator.py).

Install

git clone https://github.com/youyupei/SLSim
cd SLSim

Example code

1. Generate perfect reads (template.fa)

python3 bin/perfect_read_generator.py -r reference_transcript.fa -i bc_umi.csv -o template.fa

The output from perfect_read_generator.py is a FASTA file, each entry is an error-free read with the true barcode followed by a index as read id.

2. Simulate nanopore errors and create fastq file

python3 bin/error_simulator.py -t template.fa

The output from error_simulator.py is a FASTQ file, each entry is a error-containing read with per-base quality score. The read id can be matched to those in the FASTA file generated from perfect_read_generator.py.