Building transcriptome indexes

Introduction

Txseq uses Hisat2 for alignment-based gene expression quantification and Salmon for quasi-alignment based gene expression quantification. To use these tools, it is first necessary to build method-specific transcriptomes indexes.

As with preparation of the sanitised genome sequences and gene annotations, it is recommended to build transcriptome indexes in a central location for use in multiple projects.

Txseq has dedicated pipelines for building Salmon and Hisat2 indexes which can be used as follows.

Note

If you are using the KIR BMRC workspace, Salmon and Hisat index built with txseq can be found in the "/well/kir/projects/mirror/txseq/" directory.

Building a Salmon Index

In a suitably named directory, load the required modules and obtain the salmon_index configuration file with the following commands:

    module purge
    unset PYTHONPATH
    source ./txseq_smk_venv/bin/activate
    module load Salmon/1.11.4-GCC-12.3.0
    mkdir salmon_index
    cd salmon_index
    txseq salmon_index config
    emacs salmon_index.yaml 

After editing the yaml to point to the location of the "txseq.genome.fa.gz" and "txseq.transcript.fa.gz" files (see the "Genomes and Annotations" section), and configuring the parameters as desired, the pipeline can be executed with the following command:

    txseq salmon_index full 

Building a Hisat2 Index

In a suitably named directory, loads the required modules obtain the hisat_index configuration file with the following commands:

    module purge
    unset PYTHONPATH
    source ./txseq_smk_venv/bin/activate
    module load HISAT2/2.2.1-gompi-2023a
    mkdir hisat_index
    cd hisat_index
    txseq hisat_index config
    emacs hisat_index.yaml

After editing the yaml to point to the location of the "txseq.genome.fa.gz" and "txseq.geneset.gtf.gz" files (see the "Genomes and Annotations" section), and configuring the parameters as desired, the pipeline can be executed with the following command:

    txseq hisat_index full --jobs 2