Pipeline for analyzing barcoded amplicon sequencing data with Unique molecular identifiers (UMI)
UMIErrorCorrect has been published in Clinical Chemistry.
Link to the Umierrorcorrect paper
Osterlund T., Filges S., Johansson G., Stahlberg A. UMIErrorCorrect and UMIAnalyzer: Software for Consensus Read Generation, Error Correction, and Visualization Using Unique Molecular Identifiers, Clinical Chemistry, 2022;, hvac136
To run Umierrorcorrect via Docker, see the Docker documentation.
To install the UMI-errorcorrect pipeline from source, open a terminal and type the following:
pip install umierrorcorrect
After installation, try to run the pipeline:
run_umierrorcorrect.py -h
Umi-errorcorrect runs using Python 3 and requires the following programs/libraries to be installed (if you run through docker all dependencies are already handled):
Python-libraries (should be installed automatically):
pysam (v 0.8.4 or greater)
External programs:
bwa (bwa mem command is used)
Either of gzip or pigz (parallel gzip)
Install the external programs and add them to the path.
Since the umierrorcorrect pipeline is using bwa for mapping of reads, a bwa-indexed reference genome is needed. Index the reference genome with the command bwa index -a bwtsw reference.fa.
Example syntax for running the whole pipeline:
run_umierrorcorrect.py -r1 read1.fastq.gz -r2 read2.fastq.gz -ul umi_length -sl spacer_length -r reference_fasta_file.fasta -o output_directory
The run_umierrorcorrect.py pipeline performs the following steps:
- Preprocessing of fastq files (remove the UMI and spacer and puts the UMI in the header)
- Mapping of preprocessed fastq reads to the reference genome
- Perform UMI clustering, then error correcion of each UMI cluster
- Create consensus reads (one representative read per UMI cluster written to a BAM file)
- Create a consensus output file (collapsed counts per position)
- Perform variant calling.
It is also to possible to run the pipeline step-by-step.
To see the options for each step, type the following:
preprocess.py -h
run_mapping.py -h
umi_error_correct.py -h
get_consensus_statistics.py -h
call_variants.py -h
filter_bam.py -h
filter_cons.py -h