De novo Assembly

Overview

After trimming and quality checking sequencing reads, the next step is to reconstruct the bacterial genome from short sequencing reads. This process is called genome assembly.

In this lesson, we will use trimmed paired-end reads to perform de novo genome assembly. De novo assembly means that we assemble the genome without using a reference genome.

Learning objectives

By the end of this lesson, you should be able to:

  • explain what genome assembly means
  • understand the difference between raw reads and contigs
  • run a bacterial genome assembly using SPAdes
  • identify the main assembly output files
  • organize assembly outputs into a clean directory structure
  • prepare assembly files for downstream quality control

Activate the assembly environment

If you are continuing from the previous trimming lesson, you may still be in the qc environment. For assembly, we need a different environment named ’assembly. First we will deactivate qc environment to safely exit.

conda deactivate

It will return to base environment

(base) genomevm@genome-clone-vm2:~tanzim$

Now activate assembly environment:

conda activate assembly
(assembly) genomevm@genome-clone-vm2:~tanzim$

Check that spades.py is available:

spades.py --version
(assembly) genomevm@genome-clone-vm2:~tanzim$ 
SPAdes genome assembler v4.2.0

If the command works, you should see the installed SPAdes version.

Let’s check for manuals quickly.

spades.py --help

Input files

For genome assembly, we will use the paired trimmed reads generated in the previous lesson.

trimmed_reads/S1_trim_R1.fastq.gz
trimmed_reads/S1_trim_R2.fastq.gz

Check that the files are present:

ls trimmed_reads

You should see:

S1_trim_R1.fastq.gz
S1_trim_R2.fastq.gz
NoteImportant

For paired-end assembly, use the paired trimmed reads:

S1_trim_R1.fastq.gz
S1_trim_R2.fastq.gz

Do not use the unpaired files unless your instructor specifically asks you to include them.

What is genome assembly?

Genome assembly is the process of joining sequencing reads together to create longer sequences.

These longer sequences are called contigs.

Short reads  →  contigs  →  draft genome assembly

Because bacterial genomes are usually a few million base pairs long, short reads cannot cover the whole genome in one piece. Instead, the assembler reconstructs the genome as multiple contigs.

Create an assembly output folder

Create a folder for assembly results:

mkdir -p assembly

Run SPAdes

Run SPAdes using the trimmed paired-end reads:

spades.py \
  -1 trimmed_reads/S1_trim_R1.fastq.gz \
  -2 trimmed_reads/S1_trim_R2.fastq.gz \
  -o assembly/S1_spades \
  -t 4

Understanding the command

Option Meaning
spades.py Runs the SPAdes assembler
-1 Forward paired-end reads
-2 Reverse paired-end reads
-o Output directory
-t 4 Use 4 CPU threads
TipTip

If your computer or server has more available CPUs, you can increase the number after -t.

For example:

-t 8

Check assembly outputs

After SPAdes finishes, list the output directory:

ls assembly/S1_spades

You should see files and folders such as:

contigs.fasta
scaffolds.fasta
assembly_graph.fastg
spades.log

The most important output for this training is usually:

assembly/S1_spades/contigs.fasta

Important SPAdes output files

File Description
contigs.fasta Main assembled contig file
scaffolds.fasta Scaffolded assembly file
spades.log Log file from the assembly run
assembly_graph.fastg Assembly graph file

For most downstream bacterial genome analysis in this course, we will use:

contigs.fasta

Inspect the assembly file

A FASTA file contains sequence names and nucleotide sequences.

View the first few lines of the assembly:

head assembly/S1_spades/contigs.fasta

A FASTA file looks like this:

>NODE_1_length_355368_cov_93.924892
TATGAGGTGATGATATCACCTCTAACTAAGTGACC....
>NODE_2_length_343701_cov_97.093919
GATTCCATCCCGAACTCAGAAGTGAAACGAAACAG....

Each sequence starts with a header line beginning with >.

Count the number of contigs

To count the number of contigs in the assembly:

grep -c "^>" assembly/S1_spades/contigs.fasta

This counts the number of FASTA headers.

270
NoteExercise

Does your assembly contain one contig or multiple contigs?

Most short-read bacterial genome assemblies contain multiple contigs. A smaller number of contigs usually suggests a more continuous assembly, but quality should be checked using assembly QC tools.

Copy the final assembly file

To make downstream analysis easier, create a folder for final assemblies:

mkdir -p assemblies

Copy the SPAdes contig file into this folder with a clear sample name:

cp assembly/S1_spades/contigs.fasta assemblies/S1_contigs.fasta

Check the copied file:

ls assemblies

You should see:

S1_contigs.fasta

Open the downloaded file using a text editor such as Notepad++, VS Code, or another FASTA viewer.

Use:

open S1_contigs.fasta

Use:

xdg-open S1_contigs.fasta

Final directory structure

At the end of this lesson, your directory should look like this:

.
├── raw_reads
│   ├── S1_R1.fastq.gz
│   └── S1_R2.fastq.gz
├── trimmed_reads
│   ├── S1_trim_R1.fastq.gz
│   └── S1_trim_R2.fastq.gz
├── qc_reports
│   ├── S1_trim_R1_fastqc.html
│   ├── S1_trim_R1_fastqc.zip
│   ├── S1_trim_R2_fastqc.html
│   └── S1_trim_R2_fastqc.zip
├── assembly
│   └── S1_spades
│       ├── contigs.fasta
│       ├── scaffolds.fasta
│       └── spades.log
└── assemblies
    └── S1_contigs.fasta

Key points

NoteImportant
  • Genome assembly reconstructs longer genome sequences from short sequencing reads.
  • SPAdes is commonly used for bacterial de novo genome assembly.
  • The main output file is usually contigs.fasta.
  • Rename and copy the final contig file to a separate assemblies folder.
  • The assembly file will be used in the next lesson for assembly quality control.

Next step

In the next lesson, we will assess the quality of the assembled genome using assembly QC tools.

Back to top