apscale

Advanced Pipeline for Simple yet Comprehensive AnaLysEs of DNA metabarcoding data

- apscale

- apscale-nanopore

apscale-nanopore

Introduction

Apscale-nanopore is a modified version of the metabarcoding pipeline apscale and is used for the processing of Oxford Nanopore data.

Programs used:

• cutadapt
• vsearch
• swarm
• blast+ (blastn module)

Input:

• Non-demultiplexed Nanopore sequence data in .fastq format.
• Demultiplexed Nanopore sequence data in .fastq format.

Output:

• read table, taxonomy table, log files, report

Installation

Apscale-nanopore can be installed on all common operating systems (Windows, Linux, MacOS). Apscale-nanopore requires Python 3.10 or higher and can be easily installed via pip in any command line:

pip install apscale_nanopore

To update apscale-blast run:

pip install --upgrade apscale_nanopore

The easiest installation option is the Conda apscale environment. This way, all dependencies will automatically be installed.

Then activate the conda environment.

conda activate apscale

Create project

First, create a new project:

apscale_nanopore create -p PATH/TO/PROJECT

A new project will be created. Follow the instructions and fill out the settings file accordingly.

/YOUR_PROJECT_PATH/My_new_project/
├───1_raw_data
│   └───data
├───2_index_demultiplexing
│   └───data
├───3_primer_trimming
│   └───data
├───4_quality_filtering
│   └───data
├───5_clustering_denoising
│   └───data
├───6_read_table
│   └───data
├───7_taxonomic_assignment
│   └───data
├───8_nanopore_report
My_new_project_settings.xlsx

Settings file

Sample index and primer combinations (Example)

Forward index 5'-3'	Forward primer 5'-3'	Reverse index 5'-3'	Reverse primer 5'-3'	ID
AGAACGACTTCCATACTCGTGTGA	RGCHTTYCCHCGWATAAAYAAYATAAG	AGAACGACTTCCATACTCGTGTGA	GRGGRTAWACWGTTCAWCCWGTNCC	Sample_1
AACGAGTCTCTTGGGACCCATAGA	RGCHTTYCCHCGWATAAAYAAYATAAG	AACGAGTCTCTTGGGACCCATAGA	GRGGRTAWACWGTTCAWCCWGTNCC	Sample_2
AGGTCTACCTCGCTAACACCACTG	RGCHTTYCCHCGWATAAAYAAYATAAG	AGGTCTACCTCGCTAACACCACTG	GRGGRTAWACWGTTCAWCCWGTNCC	Sample_3
CGTCAACTGACAGTGGTTCGTACT	RGCHTTYCCHCGWATAAAYAAYATAAG	CGTCAACTGACAGTGGTTCGTACT	GRGGRTAWACWGTTCAWCCWGTNCC	Sample_4
ACCCTCCAGGAAAGTACCTCTGAT	RGCHTTYCCHCGWATAAAYAAYATAAG	ACCCTCCAGGAAAGTACCTCTGAT	GRGGRTAWACWGTTCAWCCWGTNCC	Sample_5
CCAAACCCAACAACCTAGATAGGC	RGCHTTYCCHCGWATAAAYAAYATAAG	CCAAACCCAACAACCTAGATAGGC	GRGGRTAWACWGTTCAWCCWGTNCC	Sample_6

Apscale-nanopore settings (Example)

Step	Category	Variable	Comment
General	cpu count	7	Number of cores to use
demultiplexing (index)	allowed errors index	3	Allowed errors during index demultiplexing
primer trimming	allowed errors primer	4	Allowed errors during primer trimming
quality filtering	minimum length	54	Reads below this length will be discarded
quality filtering	maximum length	74	Reads above this length will be discarded
quality filtering	minimum quality	20	Reads below this average PHRED quality score will be discarded
clustering/denoising	mode	denoised OTUs	Choose clustering/denoising algorithm
clustering/denoising	percid	0.97	Vsearch clustering percentage identity
clustering/denoising	alpha	1	Vsearch denoising alpha value
clustering/denoising	d	1	Swarm's d value
read table	minimum reads	10	Discard reads below this threshold
taxonomic assignment	apscale blast	Yes	Run APSCALE megablast (yes or no)
taxonomic assignment	apscale db	...	Path to local database

Run apscale-nanopore

Apscale-nanopore operates in four different ways:

1) Raw-data processing of non-demultiplexed data

• Copy your non-demultiplexed .fastq(.gz) files to the "1_raw_data/data" folder.

apscale_nanopore run -p PATH/TO/PROJECT

• Apscale-nanopore will demultiplex all your files according to the demultiplexing sheet.

2) Raw-data processing of demultiplexed data

• Copy your demultiplexed .fastq(.gz) files to the "1_raw_data/data" folder.

apscale_nanopore run -p PATH/TO/PROJECT -sd

• Apscale-nanopore will skip the demultiplexing and immediately start with the raw-data processing.
• Important: Enter the primer sequences (5'-3') in the first row of the demultiplexing sheet. The index columns can be left blank.

3) Live raw-data processing of non-demultiplexed data

• Output your non-demultiplexed .fastq(.gz) files to the "1_raw_data/data" folder during sequencing.
• Apscale-nanopore will automatically scan the folder for incoming files and automatically process them.
• Press Ctrl+C to interupt the live-calling.

apscale_nanopore run -p PATH/TO/PROJECT -l

• Apscale-nanopore will demultiplex all your files according to the demultiplexing sheet.

4) Live raw-data processing of demultiplexed data

• Output your demultiplexed .fastq(.gz) files to the "1_raw_data/data" folder during sequencing.
• Apscale-nanopore will automatically scan the folder for incoming files and automatically process them.
• Press Ctrl+C to interupt the live-calling.

apscale_nanopore run -p PATH/TO/PROJECT -l -sd

• Apscale-nanopore will skip the demultiplexing and immediately start with the raw-data processing.
• Important: Enter the primer sequences (5'-3') in the first row of the demultiplexing sheet. The index columns can be left blank.

Run individual steps

• Apscale can run individual steps (-step X) or all steps after a specific module (-steps X).

Step indices:

• 1 = Index demultiplexing
• 2 = Primer trimming
• 3 = Quality filtering
• 4 = Clustering/denoising
• 5 = Read table
• 6 = Taxonomic assignment

Example: Run "clustering/denoising"

apscale_nanopore run -p PATH/TO/PROJECT -step 4

Example: Run all steps after the "quality filtering":

apscale_nanopore run -p PATH/TO/PROJECT -steps 3

Quality control

A quality control can be conducted for all fastq files. Simply run:

apscale_nanopore qc -p PATH/TO/PROJECT

Bioinformatics Workflow Overview

1) Demultiplexing

Tool: cutadapt Settings: Allowed errors (default=3)

Demultiplex raw sequencing reads based on barcode sequences to generate sample-specific FASTQ files.

---

2) Primer Trimming

Tool: cutadapt Settings: Allowed errors (default=4)

Remove primer sequences from demultiplexed reads to retain only target regions.

---

3) Quality Filtering

Tools: python, vsearch Settings: Min. mean Q-Score (default=20), Min. and max. length (fragment-specific)

Filter reads based on:

• Mean PHRED quality score
• Minimum and maximum fragment length

This step ensures only high-quality reads are retained for downstream processing.

---

4) Clustering / Denoising

Tool: vsearch
Settings: d (default=1), percentage identity (default=0.97), alpha (default=1)

Choose from the following processing strategies:

• Swarm denoising: Local clustering using the Swarm algorithm for fine-scale resolution.
• Swarm OTUs: Swarm denoising followed by similarity clustering.
• ESV denoising: Error-correction to obtain Exact Sequence Variants.
• Denoised OTUs: Denoising followed by similarity clustering.

---

5) Read Table Construction and Filtering

Tool: python
Settings: minimum reads (default=10)

Construct an abundance table (ESVs/OTUs × samples).

Apply a minimum read threshold to remove low-abundance features.

---

6) Taxonomic Assignment

Tool: BLASTn via apscale-blast

Settings: Apscale-blast database

Assign taxonomy to representative sequences using a local reference database.

---

7) Quality Control and Reporting

Tool: python

Generate summary statistics and visual diagnostics.