Installation

primer-finder is a Python package with no Python dependency; what it needs are the programs it runs: KMC, SKESA or SPAdes (or neither, with -a unitig), minimap2 and BLAST. Python 3.10 or later.

From bioconda (recommended)

Not available yet: the recipe is awaiting review (bioconda-recipes#70034). Until it is merged, install from PyPI into an environment that already holds the programs, or from the source code with environment.yml, which brings them.

conda create -n primer-finder -c conda-forge -c bioconda primer-finder
conda activate primer-finder
primer-finder --version

Mamba works the same way (mamba create ...) and is much faster.

From PyPI

pip install primer-finder installs the command, but not the programs it runs: KMC, SKESA or SPAdes, minimap2 and BLAST are not Python packages. Use it inside an environment that already holds them (or install them separately); otherwise prefer the bioconda package above, which brings everything.

pip install primer-finder   # or: pipx install primer-finder

From the source code

The conda environment file brings the external programs in, then pip installs primer-finder itself:

git clone https://github.com/duceppemo/primer-finder
cd primer-finder
conda env create -f environment.yml
conda activate primer-finder
pip install .

Without installing anything, python primer_finder.py ... works from a clone and takes the same options as primer-finder.

The programs it runs

Program	Used for	conda package
kmc, kmc_tools	counting and subtracting kmers	kmc >=3.2
skesa	assembling the kmers (-a skesa, the default)	skesa >=2.4
spades.py	assembling the kmers (-a spades)	spades >=3.15
minimap2	mapping the contigs to an exclusion genome	minimap2 >=2.24
blastn, makeblastdb	checking the contigs against every genome	blast >=2.14
primer3_core	designing assays on the regions (primer-finder design)	primer3 >=2.6

A missing program is reported before anything runs, with the conda package that provides it. Only the assembler you ask for has to be installed, and -a unitig asks for none.

insilicoPCR is not installed with primer-finder and is not needed to run it. It is a separate program, and primer-finder design --insilico-pcr points at an extracted portable release of it (which brings its own Java, BBMap and BLAST+). See Designing assays.

Checking the installation

The repository holds a small example: simulated genomes whose specific region is known, which the whole pipeline must find. It takes a few seconds.

# In a clone
bash example/run_example.sh

# Or, with primer-finder installed from conda
curl -sL https://github.com/duceppemo/primer-finder/archive/refs/tags/v1.8.0.tar.gz | tar -xz --strip-components=1 primer-finder-1.8.0/example
bash example/run_example.sh

It ends with All checks passed. See Example for what it checks.

Upgrading from the 2022 scripts

The two scripts became a package, and the old command lines still work:

Before	Now
python primer_finder.py -i incl/ -e excl/ -o out/	primer-finder -i incl/ -e excl/ -o out/ (or the same primer_finder.py command)
python IDT_results_converter.py sheet.xlsx out.fasta prefix	primer-finder idt sheet.xlsx out.fasta prefix

The options are unchanged. What moved is where the results are written: final_kmers.fasta is still at the top of the output folder, and the intermediate files are now in numbered subfolders (see Outputs). pysam, Biopython, pandas, psutil, pyahocorasick and xlrd are no longer needed, and neither are bowtie2 and samtools; requirements.txt is gone, replaced by environment.yml.