KrakenParser: Convert Kraken2 Reports to CSV

Overview

KrakenParser is a collection of scripts designed to process Kraken2 reports and convert them into CSV format. This pipeline extracts taxonomic abundance data at six levels:

• Phylum
• Class
• Order
• Family
• Genus
• Species

You can run the entire pipeline with a single command, or use the scripts individually depending on your needs.

🔗 Please visit KrakenParser wiki page

Output example

Total abundance output

counts_phylum.csv parsed from 9 kraken2 reports of metagenomic samples using KrakenParser:

Sample_id,Calditrichota,Caldisericota,Thermosulfidibacterota,Elusimicrobiota,Candidatus Fervidibacterota,Lentisphaerota,Kiritimatiellota,Vulcanimicrobiota,Thermodesulfobiota,Atribacterota,Dictyoglomota,Nitrospinota,Chrysiogenota,Coprothermobacterota,Aquificota,Thermotogota,Bdellovibrionota,Nitrospirota,Deferribacterota,Synergistota,Myxococcota,Acidobacteriota,Candidatus Bipolaricaulota,Candidatus Saccharibacteria,Candidatus Absconditabacteria,Fusobacteriota,Spirochaetota,Candidatus Omnitrophota,Chlamydiota,Verrucomicrobiota,Planctomycetota,Thermodesulfobacteriota,Campylobacterota,Candidatus Cloacimonadota,Fibrobacterota,Gemmatimonadota,Balneolota,Rhodothermota,Ignavibacteriota,Chlorobiota,Bacteroidota,Deinococcota,Thermomicrobiota,Armatimonadota,Chloroflexota,Cyanobacteriota,Mycoplasmatota,Actinomycetota,Bacillota,Pseudomonadota,Heterolobosea,Parabasalia,Fornicata,Evosea,Bacillariophyta,Cercozoa,Euglenozoa,Apicomplexa,Microsporidia,Basidiomycota,Ascomycota,Nanoarchaeota,Candidatus Micrarchaeota,Candidatus Thermoplasmatota,Candidatus Lokiarchaeota,Nitrososphaerota,Euryarchaeota,Thermoproteota,Hofneiviricota,Artverviricota,Nucleocytoviricota,Cossaviricota,Kitrinoviricota,Negarnaviricota,Lenarviricota,Pisuviricota,Peploviricota,Uroviricota
X1,0,0,0,0,0,0,0,0,1,1,1,1,2,3,4,5,7,8,9,17,23,25,5,13,22,47,54,1,6,27,31,128,151,2,6,13,1,3,7,44,14991,7,9,11,61,414,449,3551,55304,438645,0,0,0,0,0,0,1,22,0,4,15,0,0,0,0,0,3,191,0,0,1,88,0,0,0,161,0,1241
X2,1,4,14,20,5,12,15,6,8,15,2,15,109,68,182,97,79,196,70,272,331,149,36,77,35,562,1237,21,33,129,427,1044,543,8,98,25,16,45,11,1043,41374,160,28,161,1348,1196,2709,15864,431170,2747842,22,7,301,373,134,136,107,3239,54,1151,2905,0,0,3,5,6,7,410,0,0,0,736,0,3,11,26,1,1552
...
X8,1,19,0,47,0,1,6,20,28,0,1,1,47,7,336,110,30,32,10,93,85,48,9,7,7,154,386,0,14,19,106,358,242,14,5,134,15,11,7,18,54057,106,10,24,212,340,1128,16220,567908,650264,95,4,193,402,314,300,187,4376,37,9796,8653,0,1,0,1,5,23,1778,1,1,0,1,1,4,66,30,4,1263
X9,0,3,2,16,7,1,23,12,10,9,1,2,134,40,390,289,29,372,27,81,150,90,9,88,32,287,881,14,33,60,319,1045,328,15,22,22,10,72,8,63,35301,127,15,48,412,935,2343,11500,380765,2613854,0,0,0,0,0,0,5,74,0,38,40,3,0,0,0,1,3,275,0,0,0,0,0,2,118,25,0,1675

Relative abundance output

ra_phylum.csv calculated from 9 kraken2 reports of metagenomic samples using KrakenParser:

Sample_id,taxon,rel_abund_perc
X1,Pseudomonadota,85.03558294577552
X1,Bacillota,10.72121619814011
X1,Other (<4.0%),4.243200856084384
X2,Pseudomonadota,84.28702055549813
X2,Bacillota,13.225663867469137
X2,Other (<4.0%),2.487315577032736
...
X8,Pseudomonadota,49.25373021277305
X8,Bacillota,43.01574040339849
X8,Bacteroidota,4.094504530639667
X8,Other (<4.0%),3.6360248531887933
X9,Pseudomonadota,85.62839981589192
X9,Bacillota,12.473649123439218
X9,Other (<4.0%),1.8979510606688494

α-diversity output

alpha_div.csv calculated from 9 kraken2 reports of metagenomic samples using KrakenParser:

Sample,Shannon,Pielou,Chao1
X1,3.911345447107001,0.5269245043289149,2274.533185840708
X2,3.9944130792536563,0.4906424221265042,4155.0
...
X8,3.442077115880119,0.42753293021330063,4177.251358695652
X9,4.033664950188261,0.5050385978575492,3492.16

β-diversity output

beta_div_bray.csv calculated from 9 kraken2 reports of metagenomic samples using KrakenParser:

,X1,X2,...,X8,X9
X1,0.0,0.398,...,0.61,0.353
X2,0.398,0.0,...,0.723,0.388
...
X8,0.61,0.723,...,0.0,0.665
X9,0.353,0.388,...,0.665,0.0

beta_div_jaccard.csv calculated from 9 kraken2 reports of metagenomic samples using KrakenParser:

,X1,X2,...,X8,X9
X1,0.0,0.7073170731707317,...,0.8223938223938224,0.7232472324723247
X2,0.7073170731707317,0.0,...,0.835016835016835,0.7352941176470589
...
X8,0.8223938223938224,0.835016835016835,...,0.0,0.8066914498141264
X9,0.7232472324723247,0.7352941176470589,...,0.8066914498141264,0.0

Visualization examples gallery

Stacked Barplot	Streamgraph

Stacked Barplot + Streamgraph	Clustermap

Quick Start (Full Pipeline)

To run the full pipeline, use the following command:

KrakenParser --complete -i data/kreports -o results/
#Having troubles? Run KrakenParser --complete -h

For reproducible β-diversity (rarefaction is stochastic by default):

KrakenParser -i data/kreports -o results/ -s 42

This will:

1. Convert Kraken2 reports to MPA format
2. Combine MPA files into a single file
3. Extract taxonomic levels into separate text files
4. Process extracted text files
5. Convert them into CSV format
6. Calculate relative abundance
7. Calculate α & β-diversities

Installation

pip install krakenparser

Before Visualization: Grouping Low-Abundance Taxa

The full pipeline automatically calculates relative abundance. Before passing data to visualization, it is strongly recommended to re-run --relabund with the -O flag — this collapses all taxa below the chosen threshold into a single "Other" group, producing much cleaner and more readable plots.

KrakenParser --relabund -i data/counts/counts_species.csv -o data/rel_abund/ra_species.csv -O 4

This groups every taxon with relative abundance < 4 % into Other (<4.0%). Adjust the threshold to your data.

Note: The pipeline-generated rel_abund/ra_*.csv files (no -O) preserve the full unfiltered data — use them for statistical analysis. Use the -O variant specifically for visualization.

---

Using Individual Modules (Advanced)

Each step of the pipeline can also be run individually. This is useful for re-running a single step, debugging, or integrating KrakenParser into a custom workflow.

Step 1: Convert Kraken2 Reports to MPA Format

# Batch mode (directory)
KrakenParser --kreport2mpa -i data/kreports -o data/intermediate/mpa
# Single file
KrakenParser --kreport2mpa -r data/kreports/sample.kreport -o data/intermediate/mpa/sample.MPA.TXT
#Having troubles? Run KrakenParser --kreport2mpa -h

Converts Kraken2 .kreport files into MPA format.

Step 2: Combine MPA Files

KrakenParser --combine_mpa -i data/intermediate/mpa/* -o data/intermediate/COMBINED.txt
#Having troubles? Run KrakenParser --combine_mpa -h

Merges multiple MPA files into a single combined table.

Step 3: Extract Taxonomic Levels

KrakenParser --deconstruct -i data/intermediate/COMBINED.txt -o data/intermediate
#Having troubles? Run KrakenParser --deconstruct -h

By default, human-related taxa (Homo sapiens, Hominidae, Primates, Mammalia, Chordata) are removed. To keep them:

KrakenParser --deconstruct -i data/intermediate/COMBINED.txt -o data/intermediate --keep-human

To inspect the Viruses domain separately:

KrakenParser --deconstruct_viruses -i data/intermediate/COMBINED.txt -o data/counts_viruses
#Having troubles? Run KrakenParser --deconstruct_viruses -h

Step 4: Process Extracted Taxonomic Data

KrakenParser --process -i data/intermediate/COMBINED.txt -o data/intermediate/txt/counts_phylum.txt
#Having troubles? Run KrakenParser --process -h

Repeat on other 5 taxonomical levels (class, order, family, genus, species) or wrap up KrakenParser --process in a loop.

Cleans up taxonomic names: removes prefixes (s__, g__, etc.) and replaces underscores with spaces.

Step 5: Convert TXT to CSV

KrakenParser --txt2csv -i data/intermediate/txt/counts_phylum.txt -o data/counts/counts_phylum.csv
#Having troubles? Run KrakenParser --txt2csv -h

Repeat on other 5 taxonomical levels or wrap in a loop. Transposes data so that sample names become rows.

Step 6: Calculate Relative Abundance

KrakenParser --relabund -i data/counts/counts_phylum.csv -o data/rel_abund/ra_phylum.csv
#Having troubles? Run KrakenParser --relabund -h

Repeat on other 5 taxonomical levels or wrap in a loop.

With "Other" grouping:

KrakenParser --relabund -i data/counts/counts_phylum.csv -o data/rel_abund/ra_phylum.csv -O 3.5

Groups all taxa with abundance < 3.5 % into Other (<3.5%).

Step 7: Calculate α & β-Diversities

KrakenParser --diversity -i data/counts/counts_species.csv -o data/diversity
#Having troubles? Run KrakenParser --diversity -h

With a custom rarefaction depth:

KrakenParser --diversity -i data/counts/counts_species.csv -o data/diversity -d 750

For reproducible results (rarefaction uses random subsampling — fix the seed to get the same matrix every run):

KrakenParser --diversity -i data/counts/counts_species.csv -o data/diversity -s 42

---

Arguments Breakdown

--complete (Full Pipeline)

• Requires -i: path to the Kraken2 reports directory (e.g., data/kreports).
• Optional -o: output directory (default: parent of -i).
• Optional --keep-human: retain human-related taxa (default: filtered out).
• Optional -s INT: random seed for reproducible β-diversity rarefaction (default: random).

--kreport2mpa (Step 1)

• Batch mode: -i DIR -o DIR — converts all files in a directory.
• Single-file mode: -r FILE -o FILE.

--combine_mpa (Step 2)

• -i FILE [FILE ...]: one or more MPA files.
• -o FILE: output merged table.

--deconstruct & --deconstruct_viruses (Step 3)

• Extracts phylum, class, order, family, genus, species into separate text files.
• --deconstruct removes human-related reads by default; use --keep-human to retain them.
• --deconstruct_viruses extracts only the Viruses domain.

--process (Step 4)

• Removes prefixes (s__, g__, etc.), replaces underscores with spaces.
• -i: COMBINED.txt (source for sample-name header); -o: target txt file.

--txt2csv (Step 5)

• Transposes a processed txt file into a CSV with sample names as rows.

--relabund (Step 6)

• Calculates relative abundance from a total-counts CSV.
• -O FLOAT: group taxa below FLOAT % into Other (<FLOAT%).

--diversity (Step 7)

• Shannon, Pielou & Chao1 for α-diversity.
• Bray-Curtis & Jaccard for β-diversity.
• -d INT: rarefaction depth for β-diversity (default: 1000).
• -s INT: random seed for reproducible rarefaction (default: random — results vary between runs).

Example Output Structure

After running the full pipeline, the output directory will look like this:

results/
├─ counts/                 # Total abundance CSV output
│  ├─ counts_species.csv
│  ├─ counts_genus.csv
│  ├─ ...
│  └─ counts_phylum.csv
├─ rel_abund/              # Relative abundance CSV output
│  ├─ ra_species.csv
│  ├─ ra_genus.csv
│  ├─ ...
│  └─ ra_phylum.csv
├─ diversity/              # Diversity metrics
│  ├─ alpha_div.csv
│  ├─ beta_div_bray.csv
│  └─ beta_div_jaccard.csv
└─ intermediate/           # Intermediate files
   ├─ mpa/                 # Converted MPA files
   │  ├─ {sample}.txt
   │  ├─ ...
   ├─ COMBINED.txt         # Merged MPA table
   └─ txt/                 # Extracted taxonomic levels in TXT
      ├─ counts_species.txt
      ├─ counts_genus.txt
      ├─ ...
      └─ counts_phylum.txt

Conclusion

KrakenParser provides a simple and automated way to convert Kraken2 reports into usable CSV files for downstream analysis. You can run the full pipeline with a single command or use individual scripts as needed.

For any issues or feature requests, feel free to open an issue on GitHub!

🚀 Happy analyzing!