Introduction

genomic-medicine-sweden/metaval is a bioinformatics pipeline for post-processing the results of nf-core/taxprofiler. It verifies the taxa classified by the nf-core/taxprofiler pipeline using Nanopore and Illumina shotgun metagenomic sequencing data. At the moment, genomic-medicine-sweden/metaval only verifies the classification results from three taxonomic classifiers Kraken2, Centrifuge and DIAMOND.

The pipeline, constructed using the nf-core template, utilizing Docker/Singularity containers for easy installation and reproducible results. The implementation follows Nextflow DSL2, employing one container per process for simplified maintenance and dependency management. Processes are sourced from nf-core/modules for broader accessibility within the Nextflow community.

Pipeline summary

Green Workflow - Pathogen Screening

This workflow is activated by enabling the --perform_screen_pathogens option.

1. Map reads to pathogen genomes

   • Map reads to a predefined list of viral pathogen genomes using Bowtie2 for Illumina reads or minimap2 for Nanopore reads. This step checks for the presence of known pathogens in the sample.

2. Call consensus

   • This step calls consensus sequences for reads mapped to pathogen genomes using either samtools or medaka, depending on the read type. samtools can be used to generate consensus sequences for both Illumina and Nanopore reads, while medaka is typically used for Nanopore reads. The generated consensus sequence will be used as input for BLAST.

3. BLAST for pathogen identification

   • Use BLAST to identify the closest reference genomes for the target reads. There are two options: BLASTx using DIAMOND based on the protein database, and BLASTn based on the nucleotide database.

4. Extract target reads

   • From the mapped reads, extract the target reads that match the predefined viral pathogens based on the result of BLAST.

5. Visualisation using IGV

   • Visualize the extracted reads using IGV (Integrative Genomics Viewer) to provide a graphical representation for detailed analysis.

6. Perform quality check

   • Conduct quality checks on the target reads using FastQC and MultiQC to ensure data quality and reliability.

Orange Workflow - Verify Identified Viruses

This workflow is activated by enabling the --perform_extract_reads option and disabling the --taxid.

1. Decontamination

   • Filter the output files from metagenomics classifiers like Kraken2, Centrifuge, or DIAMOND to remove false positives and background contaminations. This step compares results to the negative control and identifies likely present species based on user-defined thresholds.

2. Extract viral TaxIDs

   • Extract viral TaxIDs predicted by taxonomic classification tools such as Kraken2, Centrifuge, and DIAMOND.

3. Extract reads

   • Extract the reads classified as viruses based on a list of identified TaxIDs.

4. de-novo assembly

   • This step performs de novo assembly for TaxIDs with a number of reads exceeding params.min_read_counts. Spades is used for Illumina reads, and Flye is used for Nanopore reads. The resulting contig files will be used as input for BLAST.

5. BLAST

   • Use BLAST to identify the closest reference genomes for the target reads. There are two options: BLASTx using DIAMOND based on the protein database, and BLASTn based on the nucleotide database.

6. Mapping

   • Map the reads of TaxIDs to the closest reference genomes identified by BLAST. Use Bowtie2 for Illumina reads and minimap2 for Nanopore reads.

7. Visualisation using IGV

   • Visualize the mapped reads using IGV.

8. Perform quality check

   • Conduct quality checks on the classified reads using FastQC and MultiQC to ensure the accuracy of the data.

Blue Workflow - Verify User-Defined TaxIDs

This workflow is activated by enabling the --perform_extract_reads option and the --taxid option, allowing users to define a list of TaxIDs. It is not limited to viral TaxIDs and can include bacteria, fungi, archaea, parasites, or plasmids.

All steps are the same as the Orange Workflow except using user-defined TaxIDs instead of extracting predefined viral TaxIDs.

Usage

[!NOTE] If you are new to Nextflow and nf-core, please refer to this page on how to set-up Nextflow. Make sure to test your setup with -profile test before running the workflow on actual data.

First, prepare a samplesheet with your input data that looks as follows:

samplesheet.csv:

csv sample,run_accession,instrument_platform,fastq_1,fastq_2,kraken2_report,kraken2_result,kraken2_taxpasta,centrifuge_report,centrifuge_result,centrifuge_taxpasta,diamond,diamond_taxpasta sample1,run1,ILLUMINA,sample1.unmapped_1.fastq.gz,sample1.unmapped_2.fastq.gz,sample1.kraken2.kraken2.report.txt,sample1.kraken2.kraken2.classifiedreads.txt,kraken2_kraken2-db.tsv,sample1.centrifuge.txt,sample1.centrifuge.results.txt,centrifuge_centrifuge-db.tsv,sample1.diamond.tsv,diamond_diamond-db.tsv sample2,run1,ILLUMINA,sample2.unmapped_1.fastq.gz,sample2.unmapped_2.fastq.gz,sample2.kraken2.kraken2.report.txt,sample2.kraken2.kraken2.classifiedreads.txt,kraken2_kraken2-db.tsv,sample2.centrifuge.txt,sample2.centrifuge.results.txt,centrifuge_centrifuge-db.tsv,sample2.diamond.tsv,diamond_diamond-db.tsv

Each row represents a fastq file (single-end) or a pair of fastq files (paired end).

Now, you can run the pipeline using:

bash nextflow run genomic-medicine-sweden/metaval \ -profile <docker/singularity/.../institute> \ --input samplesheet.csv \ --outdir <OUTDIR> \ --perform_extract_reads --extract_kraken2_reads

[!WARNING] Please provide pipeline parameters via the CLI or Nextflow -params-file option. Custom config files including those provided by the -c Nextflow option can be used to provide any configuration except for parameters; see docs.

For more details and further functionality, please refer to the usage documentation.

Test data

There are three test datasets within assets/test_data/, produced by the nf-core/taxprofiler pipeline

• taxprofiler_test_data: produced by executing the test.config file within the pipeline nf-core/taxprofiler.
• taxprofiler_test_full_data: produced by executing the test_full.config file within the pipeline nf-core/taxprofiler.
• test_data_version2_subset: produced by running the data downloaded from https://www.nature.com/articles/s41598-021-83812-x

The corresponding input samplesheets are stored in assets/

• samplesheet_v1.csv:results of taxprofiler test data; limited classification results; no viruses; single-end (perform_runmerging).
• samplesheet_v2.csv:results of taxprofiler full test data; no viruses; single-end (perform_runmerging).
• samplesheet_v3.csv: with viruses; subset data from test_data_version2_subset (sample 20% of pair-end reads).

Headlines of input files

kraken2_report & centrifuge_report

csv 4.62 167021 167021 U 0 unclassified 95.38 3445908 335 R 1 root 95.36 3445179 323 R1 131567 cellular organisms 93.28 3369988 622 D 2759 Eukaryota 93.26 3369247 30 D1 33154 Opisthokonta

kraken2_result

csv C SRR13439790.3 9606 150|150 9606:4 0:18 9606:7 0:5 9606:15 0:19 9606:9 0:2 9606:13 33154:1 9606:9 0:9 9606:5 |:| 9606:26 0:1 9606:3 0:32 9606:2 0:10 9606:3 0:21 9606:17 0:1 C SRR13439790.5 9606 103|103 9606:5 0:38 9606:5 0:3 9606:8 0:2 9606:8 |:| 9606:13 0:56 C SRR13439790.7 9606 150|150 9606:60 0:4 9606:1 0:1 9606:6 0:26 9606:2 0:7 9606:9 |:| 0:5 9606:1 0:44 9606:4 0:7 9606:1 0:21 9606:20 2759:4 9606:9 C SRR13439790.8 9606 107|107 0:3 9606:23 0:3 9606:14 0:16 9606:14 |:| 9606:3 0:51 9606:11 0:8 C SRR13439790.9 9606 101|150 0:48 9606:1 0:18 |:| 0:8 9606:5 0:103

centrifuge_result

csv readID seqID taxID score 2ndBestScore hitLength queryLength numMatches SRR13439790.3 NT_187391.1 9606 1624 557 109 300 1 SRR13439790.5 NC_000022.11 9606 905 169 96 206 1 SRR13439790.7 NC_000007.14 9606 6025 961 125 300 1 SRR13439790.9 unclassified 0 0 0 0 251 1

diamond

csv SRR13439790.3 0 0 SRR13439790.3 0 0 SRR13439790.5 0 0 SRR13439790.5 0 0 SRR13439790.7 0 0

Pipeline output

For more details about the output files and reports, please refer to the output documentation.

Credits

genomic-medicine-sweden/metaval was originally written by LilyAnderssonLee.All PRs were reviewed by sofstam, with additional contributions from lokeshbio.

We thank the following people for their extensive assistance in the development of this pipeline:

Contributions and Support

If you would like to contribute to this pipeline, please see the contributing guidelines.

Citations

An extensive list of references for the tools used by the pipeline can be found in the CITATIONS.md file.

This pipeline uses code and infrastructure developed and maintained by the nf-core community, reused here under the MIT license.

The nf-core framework for community-curated bioinformatics pipelines.

Philip Ewels, Alexander Peltzer, Sven Fillinger, Harshil Patel, Johannes Alneberg, Andreas Wilm, Maxime Ulysse Garcia, Paolo Di Tommaso & Sven Nahnsen.

Nat Biotechnol. 2020 Feb 13. doi: 10.1038/s41587-020-0439-x.