LQC: Long-read RNA-seq Quality Control

The Long-read RNA-seq quality control software.

LQC is used to generate quality control summary report for mapped SAM/BAM files of long-read RNA-seq data (PacBio, Oxford Nanopore). LQC provides detailed information about the indels, mismatches and splicing sites in the BAM files, which provides a good reference for evaluation for the sequencing quality of the long-read sequencing data.

Please cite us:

Dependencies

The LQC software was developed with python3, which demands several python packages.

• python3.11+: with os, sys, argparse, re, functools, collections, multiprocessing, shutil.
• numpy: 1.10+
• pandas: 1.0+
• matplotlib: 2.0+
• pysam: 0.16+

Bootstrap 5.3.8 is used by the final html report. It is vendored (MIT license, no Popper bundle) and inlined, so the report is a single self-contained file that opens offline.

Installation

It's advised to install the software into a virtual environment.

Create virtual environment:

conda create -n lqc
conda activate lqc

Or:

virtualenv ~/.env/lqc
source ~/.env/lqc/bin/activate

From github (development)

Clone and install with uv:

git clone https://github.com/gxiaolab/LQC
cd LQC
uv sync

This creates a .venv and installs the lqc command. Run the checks and the CLI:

uvx ruff check
uv run pytest
uv run lqc -b <cs-tagged.sorted.indexed.bam> -o out

From pip

pip install lqc

Usage

LQC accepts SAM or BAM file with cs tag (generated with --cs options by minimap2), or MD tag. If the SAM/BAM file has only MD tag, a genome fasta reference file is required as well to get the splicing information. Since cs tag included the splicing site information, a genome fasta reference is not required by SAM/BAM with cs tag.

usage: lqc [-h] -b BAM_FILE [--genome-fasta GENOME_FASTA] [-o OUTPUT_DIR] [--output-cs]
           [--output-pickle] [-c [CONTIG ...]] [-t THREAD] [--log-level LOG_LEVEL] [--version]

The Long-read RNA-seq quality control software.

optional arguments:
  -h, --help            show this help message and exit
  -b BAM_FILE, --bam-file BAM_FILE
                        input bam file, with cs tags, sorted and indexed
  --genome-fasta GENOME_FASTA
                        path of genome fasta file
  -o OUTPUT_DIR, --output_dir OUTPUT_DIR
                        directory to store output files
  --output-cs           output processed cs tags
  --output-pickle       output pickle file of results
  -c [CONTIG ...], --contig [CONTIG ...]
                        contigs to be analyzed
  -t THREAD, --thread THREAD
                        threads to be used in calculation
  --log-level LOG_LEVEL
                        logging level (default INFO): [DEBUG, INFO]
  --version             show program's version number and exit

The output directory should be empty to allow the storage of output files.

Output

By default, three kinds of output files are generated by LQC: summary tables, figures, and an HTML report. With the --output-cs option, the processed cs tag information of the BAM file is also written. Similarly, with the --output-pickle option, the result objects are written to one pickle file, which can be used for further analysis.

Summary table will be stored in the table subdirectory of the output directory. Figures will be stored in the fig subdirectory of the output directory.

Processed cs tag file has six columns: read name, contig, low, high, cs mark, cs value.

The LQC will also generate a html report for checking of the BAM file qualities.

Screenshot of LQC report: