3  Taxonomic Profiling

High-quality reads obtained after preprocessing were taxonomically classified using Kraken2 to characterize the microbial composition associated with Deroceras laeve samples.

3.0.1 Custom Database Construction

A custom multi-kingdom reference database was built using Kraken2 and NCBI taxonomy resources.

The database included the following libraries:

• Bacteria
  
• Archaea
  
• Viruses
  
• Fungi
  
• Human
  
• Plant
  
• Animal

module load kraken/2.0.8-beta

kraken2-build --download-taxonomy \
  --threads 20 \
  --db multi_kingdom_db

kraken2-build --download-library bacteria --threads 20 --db multi_kingdom_db
kraken2-build --download-library viral   --threads 20 --db multi_kingdom_db
kraken2-build --download-library fungi   --threads 20 --db multi_kingdom_db
kraken2-build --download-library archaea --threads 20 --db multi_kingdom_db
kraken2-build --download-library human   --threads 20 --db multi_kingdom_db
kraken2-build --download-library plant   --threads 20 --db multi_kingdom_db
kraken2-build --download-library animal  --threads 20 --db multi_kingdom_db

kraken2-build --build \
  --db multi_kingdom_db \
  --threads 30

3.0.2 Read Classification

Filtered paired-end reads generated after Cutadapt trimming were classified against the custom database.

kraken2 --db multi_kingdom_db \
  --threads 20 \
  --paired dlaeve1_R1cut.fastq dlaeve1_R2.fastq \
  --output dlaeve1.kraken \
  --report dlaeve1.report

3.0.3 Output Reports

For each sample, Kraken2 generated:

Classification output files (*.kraken)

Taxonomic summary reports(*.report)

These reports were used for downstream visualization and abundance analyses.

3.1 Exploratory Visualization in Pavian

Kraken2 reports were explored in Pavian to inspect taxonomic assignments and generate Sankey diagrams summarizing multi-kingdom composition.

Code

pavian::runApp(port = 5000)

This step facilitated rapid inspection of kingdom-level distributions across samples.

3.2 BIOM Table Generation

Per-sample Kraken2 reports were merged into a BIOM-formatted table for downstream analyses in R.

kraken_biom.py \
  dlaeve1.report \
  --fmt json \
  -o dlaeve_taxonomy.biom

3.2.1 Transition to Downstream Analyses

The resulting BIOM table was imported into R using the phyloseq package for taxonomic filtering, abundance summaries, and graphical analyses.

These steps are detailed in the next chapter: Taxonomic Analysis.