Identification of Lineages, Sequence Types and Resistance

Profiling bacteria and viruses from sequencing data is essential for understanding the identity, diversity, and dynamics of infectious disease populations.

Typically this can involve multiple analytical layers, from broad identification of the organisms present in a sample (which can also detect potential contamination) to detailed characterization of sub-species and lineages or serotypes.

Tools like Kraken can perform the first step: classifying sequencing reads by comparing them against large reference databases using k-mer-based algorithms. This initial, high-throughput profiling allows us to quickly determine the presence of bacteria, viruses, and other microbes in a sample. These general classifiers are essential for metagenomic surveillance and assessing the microbial composition of clinical or environmental samples. As the database used by Kraken is very large, we will not be running through this analysis here, but there is a nice tutorial found here if you would like to use this tool.

After the taxonomic classification, more specialized tools can be used to provide detailed species- or strain-level information. For example, TB-Profiler characterizes Mycobacterium tuberculosis strains and predicts drug resistance mutations, while Kleborate performs similar tasks for Klebsiella pneumoniae, assessing virulence and resistance markers. Other specialized software are available for important pathogenic bacteria, such as Mykrobe and Pathogenwatch. Additionally, platforms like Nextstrain focus on the phylogenetic and evolutionary dynamics of pathogens, enabling real-time tracking of how bacterial and viral lineages spread and diversify globally.

Here, we will use two of these specialised tools to profile some of our samples.

The data used for this exercise will be:

• Test3_R1.fastq.gz and Test3_R2.fastq.gz – Paired-end 250bp sequence data from Mycobacterium tuberculosis, collected from sputum and cultured before DNA library preparation.

• Test5_R1.fastq.gz and Test5_R2.fastq.gz – Paired-end 150bp sequence data from Mycobacterium tuberculosis, collected from Peru.

• assembly.fasta – The de novo assembled Klebsiella pneumoniae sample, collected in Taiwan, from practical session 1.

Lineage and antimicrobial resistance calling of Mycobacterium tuberculosis:

Here, we will use TB-Profiler which is a specialised bioinformatics tool that analyzes Mycobacterium tuberculosis sequencing data to predict drug resistance and lineage. It rapidly identifies resistance-associated and lineage-specific mutations from raw reads or alignments (in the BAM format) by comparing variation in assemblies to a database of known mutations.

Use the following code to profile the Test3 M.tb sample using the basic TB-Profiler commands:

tb-profiler profile -1 Test3_R1.fastq.gz -2 Test3_R2.fastq.gz --platform illumina --prefix Test3 --threads 2

Where the options specifed are:

Exercise: Re-run TB-Profiler with the sequencing data for sample Test5.

The files produced by TB-profiler are in the JSON format and saved in a folder called "results".

We can collate the profiling results in the JSON format for all samples into a readable .txt file using the following command:

tb-profiler collate

Open the resulting text file (it will be called tbprofiler.txt). Which lineages and sub-lineages do the TB samples belong? Is there are drug resistance predicted in the samples?

Sequence typing and antimicrobial resistance profiling of Klebsiella pneumoniae:

Kleborate can be used to profile Klebsiella species, particularly Klebsiella pneumoniae, from whole-genome sequencing data. It identifies key features such as species assignment, antimicrobial resistance genes, virulence loci, and sequence types. Escherichia species can also be profiled with Kleborate.

Unlike TB-Profiler, Kleborate will require an assembled genome in the FASTA format as the input.

We will use the de novo assembly from practical session 1 to run this analysis, which should be found in the Kleb1 folder.

Kleborate can be run using the following command:

cd Kleb1/ ## Navigate to the assembly

kleborate -a assembly.fasta -o Kleb1_kleborate -p kpsc --trim_headers 

Where the options specifed are:

Open the output file in the Kleb1_kleborate/ folder. What sequence type is our sample? Do we have any drug resistance predicted?

Next acitvity: Transmission Reconstruction