Multiply

<p align="center"><img src=".images/multiply-logo.png" width="500"></p> <p align="center">Multiplex PCR design, <i>in silico</i></p>

Overview

multiply is a command-line tool enabling the design of multiplexed PCRs for a user-specified set of target genes and/or regions. It works by first producing a set of candidate primers for each target using primer3 (multiply generate). It then computes the number of SNPs in each primer (multiply snpcheck); potential dimers betweeen pairs of primers (multiply align); and potential mispriming and off-target amplicons for each primer (multiply blast). Information from these three quality control steps is passed to a cost function, which is minimised by brute-force or with a greedy search algorithm (multiply select).

The pipeline is summarised below: <br></br>

<p align="center"><img src=".images/multiply-pipeline.png" width="700"></p>

Install

First, clone the repository to your local machine:

git clone https://github.com/JasonAHendry/multiply

Then, install the software dependencies using conda:

cd multiply
conda update conda
conda env create -f environments/run.yml
conda activate multiply-run

Finally, install multiply itself with pip:

pip install -e .

Test installation by running:

multiply

Basic usage

Part 1: Download information about your target organism.

To generate a new multiplex PCR, you first need to download the reference genome (FASTA) and information about gene locations (GFF) for your target organism. To see what organisms are available for download $^1$, type:

multiply download --available

Organisms are specified by a GenusSpecies keyword; e.g. PlasmodiumFalciparum or AnophelesGambiae. For example, one would download information about Plasmodium falciparum by running:

multiply download -g PlasmodiumFalciparum

Part 2: Create a multiplex design file.

The next step is to specify your target genes and/or region(s) with a design file. Examples of design files can be seen in the /designs folder. Genes are specified by a comma-separated list of gene identifiers (the target_ids field in the design file). Regions are specified by a separate BED file, with a fourth column that gives a unique name to each region (the bed field).

Candidate primers are generated using primer3. A collection of primer3 settings are available as JSON files in the folder settings/primer3$^2$. Multiple primer3 settings can be used in a single run of multiply by passing them as a comma-separated list to the primer3_settings field of the design file.

Part 3: Run.

After your target organism is downloaded and your design file is prepared, you can run the complete multiply pipeline with the following command:

multiply pipeline -d designs/<your-design-file.ini>

For multiplexes of moderate size (e.g. <20 targets), running the pipeline will typically take a few minutes. A directory containing results will be produced in the results directory, within sub-folder whose name is specified in your design file (the name field).

Footnotes

$1$ New organisms can be made available for download by adding them to the collection file located at genomes/collection.ini. Any organism available from PlasmoDB, EnsemblGenomes or RefSeq Genome can be added to the collection.

$2$ You can make your own primer3 settings by creating new or alterating existing JSON files in the settings/primer3 folder.

Resources

multiply uses the following external software and databases:

• primer3. Individual primer pair design. https://primer3.org/
• bedtools. Genome arithmetic. https://bedtools.readthedocs.io/en/latest/
• blastn. Local alignment search. https://blast.ncbi.nlm.nih.gov/Blast.cgi
• PlasmoDB. Plasmodium reference genome. http://plasmodb.org/plasmo/
• MalariaGEN. Plasmodium genetic diversity. https://www.malariagen.net/data
• EnsemblGenomes. Additional reference genomes. https://ensemblgenomes.org/
• RefSeq Genome. Additional reference genomes. https://www.ncbi.nlm.nih.gov/genome/

Primer dimer detection uses an alignment algorithm similar to the one described by Johnston et al. (2019).

Citation

We have a preprint available on bioRxiv.

Acknowledgements

My thanks to Nada Kubikova, who gave helpful advice on primer design for multiplex PCR; and to Dan J. Bridges, Gavin Band, Mulenga Mwenda, and Annie Forster, who tested various versions multiply.

This work was funded by the Bill and Melinda Gates Foundation (INV-003660).