fq2bam_meth

Generate BAM/CRAM output given one or more pairs of FASTQ files from bisulfite sequencing (BS-Seq). Can also optionally generate a BQSR report.

Refer to the fq2bam_meth Reference section for a detailed listing of all available options.

What is fq2bam_meth?

The tool fq2bam_meth is a fast, accurate algorithm for mapping methylated DNA sequence reads to a reference genome, performing local alignment, and producing alignment for different parts of the query sequence. It implements the baseline tool bwa-meth [^note] [^note] in a performant method using fq2bam (BWA-MEM + GATK) as a backend for processing on GPU.

Why fq2bam_meth?

fq2bam_meth is the Parabricks wrapper for bwa-meth, which will sort the output and can mark duplicates and recalibrate base quality scores in line with GATK best practices.

The Parabricks fq2bam_meth tool is capable of handling longer reads and is less sensitive to errors than other alignment algorithms. We enable fast and accurate whole-genome bisulfite sequencing (WGBS) to detect DNA-methylation at the single base pair level [^note].

Some of the advantages of using fq2bam_meth over similar tools include:

1. It is faster than many other BS-Seq alignment algorithms, making it the ideal choice for high-throughput analysis.
2. It maintains compatibility with existing CPU-based tools.

How Should I Use fq2bam_meth?

fq2bam_meth uses an accelerated version of BWA-MEM to generate BAM/CRAM output given one or more pairs of FASTQ files from BS-Seq. The user can turn-off marking of duplicates by adding the --no-markdups option. The BQSR step is only performed if the --knownSites input and --out-recal-file output options are provided; doing so will also generate a BQSR report.

Prior to running alignment, the reference genome must be converted using baseline bwa-meth. The bwa-meth indexing step produces a reference fasta file with a name formatted as fasta.bwameth.c2t. The indexing preparation step requires running bwameth.py index $REF.fasta. Baseline bwa-meth requires baseline BWA-MEM to be in the user’s path for indexing functionality. Note that indexing is a time-consuming prerequisite that should only need to be completed once per reference genome. The bwameth.py script can be found here.

Quick Start

1
# This command assumes all the inputs are in the current working directory and all the outputs go to the same place.
2
docker run --rm --gpus all --volume $(pwd):/workdir --volume $(pwd):/outputdir \
3
    --workdir /workdir \
4
    nvcr.io/nvidia/clara/clara-parabricks:4.7.1-1 \
5
    pbrun fq2bam_meth \
6
    --ref /workdir/${REFERENCE_FILE} \
7
    --in-fq /workdir/${INPUT_FASTQ_1} /workdir/${INPUT_FASTQ_2}  \
8
    --knownSites /workdir/${KNOWN_SITES_FILE} \
9
    --out-bam /outputdir/${OUTPUT_BAM} \
10
    --out-recal-file /outputdir/${OUTPUT_RECAL_FILE}

Useful Options for Performance

Suggested parameters to control host memory use:

• The parameter --bwa-normalized-queue-capacity controls the amount of batches that will be in memory to be processed. By default, the value normalized to the number of GPUs in the run is 10, corresponding to <number of GPUs> * <normalized capacity> for the total queue size. Multiple queues are used. Lowering this value would be a good strategy to reduce host memory use. However, a value that is too low may hamper performance.

Parabricks automatically uses an optimal number of streams based on the GPU’s device memory specifications (by default --bwa-nstreams auto). The user may experiment further with the --bwa-nstreams and --bwa-cpu-thread-pool parameters to potentially achieve better performance.

Additionally, for advanced performance tuning and additional control the option --bwa-primary-cpus allows for more fine-grained control of CPU threading. Each primary CPU thread drives P CPU thread pool threads as specifed with the option --bwa-cpu-thread-pool. The total number of CPU threads processing the CPU stages of alignment is the product of the --bwa-primary-cpus and --bwa-cpu-thread-pool parameters. This allows the user to control the ratio of “primary” CPU threads, which act indepently, to thread pool threads, which act in unison. Depending on the specific server configuration, input data, and GPUs available, it may be better to switch from more thread pool threads to more primary CPU threads. Changing the number of primary CPU threads may increase the CPU resources required.

• The parameter --gpuwrite uses the GPU to compress the final BAM or CRAM file for improved performance during the final stage.

  • --gpuwrite-deflate-algo can be used to control the compression ratio. See below for more details.

• The parameter --gpusort uses the GPU to sort the BAM or CRAM file.

• The option --cigar-on-gpu offloads CIGAR generation from the CPU (default) to the GPU. Consider enabling it on CPU-constrained systems, or if benchmarking with and without the option indicates the CPU is a bottleneck; it may improve overall runtime.

Compatible CPU-based bwa-meth, GATK4 Commands

The commands below are the bwa-meth-0.2.7, bwa-0.7.15, and GATK4 counterpart of the Parabricks command above. The output from these commands will be identical to the output from the above command. Refer to the Output Comparison page for comparing the results.

1
# Run bwa-meth and pipe the output to create a sorted BAM.
2
$ python bwa-meth.py \
3
    --read-group '@RG\tID:sample_rg1\tLB:lib1\tPL:bar\tSM:sample\tPU:sample_rg1' \
4
    --reference <INPUT_DIR>/${REFERENCE_FILE} <INPUT_DIR>/${INPUT_FASTQ_1} <INPUT_DIR>/${INPUT_FASTQ_2} \
5
    -t 32 -K 10000000 | \
6
  gatk SortSam \
7
    --java-options -Xmx30g \
8
    --MAX_RECORDS_IN_RAM 5000000 \
9
    -I /dev/stdin \
10
    -O cpu.bam \
11
    --SORT_ORDER coordinate
12
13
# Mark duplicates.
14
$ gatk MarkDuplicates \
15
    --java-options -Xmx30g \
16
    -I cpu.bam \
17
    -O mark_dups_cpu.bam \
18
    -M metrics.txt
19
20
# Generate a BQSR report.
21
$ gatk BaseRecalibrator \
22
    --java-options -Xmx30g \
23
    --input mark_dups_cpu.bam \
24
    --output <OUTPUT_DIR>/${OUTPUT_RECAL_FILE} \
25
    --known-sites <INPUT_DIR>/${KNOWN_SITES_FILE} \
26
    --reference <INPUT_DIR>/${REFERENCE_FILE}

Source of Mismatches

While Parabricks fq2bam_meth does not lose any accuracy in functionality when compared with BWA-mem and GATK there are several sources that can result in differences in output files.

• BWA-mem -K argument

In pair-ended mode, the chunk size specified by -K can cause small mismatches in the output BAM file. To get rid of the mismatches here, make sure to pass the same number to both baseline BWA-mem and Parabricks fq2bam_meth, e.g. -K 10000000.

• PA aux tag

  • Parabricks fq2bam_meth puts the PA tag last while BWA-mem puts it first.
  • BWA-mem rounds PA tag to 3 digits while Parabricks fq2bam_meth does not. The aux tag can be filtered by running samtools view -x <TAG>

• Unmapped reads

  • Parabricks fq2bam_meth sorts unmapped reads slightly differently than baseline GATK SortSam. Unmapped reads can be filtered with samtools by doing samtools view -F 4.

Options

• Baseline bwa-meth: https://github.com/brentp/bwa-meth/
• Bwa-meth manuscript: http://arxiv.org/abs/1401.1129
• https://doi.org/10.1038/s41587-022-01336-9