Overview

This tool identifies depletion guide candidates (sgRNAs) targeted to specific genes.

This is a generalization of https://github.com/gprezza/DASH_rRNA_depletion to allow any species and any target location(s). Typically, DASH is used for species-specific rRNA depletion, but could be used to target any gene(s).

Provided a FASTA file, a GTF file of region(s) to deplete, a PAM sequence (e.g. "NGG" for wildtype Cas9) and a guide length, CandidateFinder will:

• identify all possible PAM sequences
• search for all PAM sites on both strands of the regions to deplete
• return candidates sgRNA of specified length if their PAM site is sufficiently far from the end (sgRNA is immediately 5' of PAM site)
• filter candidates based on the thresholds for GC content, presence of heterodimer with primer, high end stabitily with primer
• off-target candidates are discarded. Off-target is defined as mapping regions outside of the target regions coupled with being adjacent to a PAM site 3' end
• compose the full oligo sequence of the candidates

Requirements

• conda must be installed
• this code repository must be downloaded and unpacked. Below, we refer to the unpacked directory as $WORKDIR. Unless otherwise specified, paths provided are assumed to be relative to $WORKDIR.

Create conda environment

Create the conda environments in the top level of $WORKDIR:

conda create -y -p env --file requirements.txt

Run tests

In the top level of $WORKDIR:

conda activate ./env

# install in editable mode
pip install -e .

Then run tests:

pytest -vv test/test.py

To run just a single test, use the -k flag, e.g.,

pytest -vv test/test.py -k test_orig_small

Usage

CandidateFinder(fasta, annotation, pam, guide_length)

Required input parameters:

  - fasta : str
    Path to genome fasta file (required)

  - annotations : str or pybedtools.BedTool
    If str, should be path to BED or GTF file that will be converted to
    pybedtools.BedTool file. This conversion will automatically handle
    1-based GTF and 0-based BED annotations to avoid off-by-one errors (required)

  - pam : str
    PAM sequence, e.g. "NGG" for wildtype Cas9 (required)

  - guide_length : int
    How long of a guide to select (required)

Optional parameters:

  - min_gc: int
    Lower threshold of acceptable GC percentage (inclusive) (optional; default=0)

  - max_gc : int
    Upper threshld of acceptable GC percentage (inclusive) (optional; default=100)

  - het_tm : int
    Degrees C threshold for Primer3's heterodimer routine (optional; default=40)

  - end_tm : int
    Degrees C threshold for Primer3's end stability routine (optional; default=30)

  - t7 : str
    T7 promoter (minus the first G) (optional; default="TTCTAATACGACTCACTATA")

  - scaffold : str
    Cas9 sgRNA scaffold (optional;
    default="GTTTTAGAGCTAGAAATAGCAAGTTAAAATAAGGCTAGTCCGTTATCAACTTGAAAAAGTGGCACCGAGTCGGTGCTTTTTT")

   - primer : str
    Reverse complement of this Cas9 primer will be added to template sgRNA oligos (optional;
    default="GGCATACTCTGCGACATCGT")

  - bowtie_index_prefix : str
    Path to directory containing existing bowtie2 index. If missing,
    index will be created (optional; default='tmp/index')

  - outfn : str
    Output files path and basename. Default is 'candidates' (optional; default='candidates')

  - plots : bool
    Generate coverage plots (optional; default=True)

  - legacy : bool
    Match output to original DASH tool (optional; default=False

Outputs

The output of CandidateFinder will be files named after the outfn parameter, which we will refer as $OUTFN. This can be a file prefix (candidates), or a path followed by a file prefix (i.e. output_dir/candidates)

• $OUTFN_oligos.csv: comma-separated file of unique full oligo sequences. They are built from the T7 promoter sequence, the candidate guide, the Cas9 sgRNA scaffold and the reverse complement of the primer.

• $OUTFN_grnas.bed: BED file of where the sgRNAs mapped to the reference genome. A sgRNA mapping to several positions in the genome will be listed at each of its locations.

• $OUTFN_grnas.fa: FASTA sequence of the sgRNAs only.

• $OUTFN_coverage.pdf: map of the sgRNAs along the reference genome