bio-genome-engineering-off-target-prediction

Version Compatibility

Reference examples tested with: pandas 2.2+

Before using code patterns, verify installed versions match. If versions differ:

• Python: pip show <package> then help(module.function) to check signatures
• CLI: <tool> --version then <tool> --help to confirm flags

If code throws ImportError, AttributeError, or TypeError, introspect the installed package and adapt the example to match the actual API rather than retrying.

Off-Target Prediction

"Check my guide RNA for off-target sites" → Search the genome for potential unintended cleavage sites allowing mismatches, then score each off-target by cutting frequency determination (CFD) to assess guide specificity.

• CLI: cas-offinder for genome-wide off-target search
• Python: CFD scoring with mismatch penalty matrices

Cas-OFFinder (CLI)

Cas-OFFinder searches genomes for potential off-target sites allowing mismatches.

# Input file format (input.txt):
# Line 1: Path to genome directory (2bit or fasta index)
# Line 2: PAM pattern (N = any, R = A/G, Y = C/T)
# Line 3+: Guide sequences with mismatch tolerance

# Example input.txt:
# /path/to/genome
# NNNNNNNNNNNNNNNNNNNNNGG
# ATCGATCGATCGATCGATCGNNN 4

# Run Cas-OFFinder
cas-offinder input.txt C output.txt  # C = use CPU
cas-offinder input.txt G output.txt  # G = use GPU (faster)

Cas-OFFinder Input Preparation

def prepare_cas_offinder_input(guides, genome_path, max_mismatches=4, pam='NGG'):
    '''Prepare Cas-OFFinder input file

    Args:
        guides: List of 20nt guide sequences
        genome_path: Path to genome directory with .2bit or indexed fasta
        max_mismatches: Maximum mismatches to search (0-6 typical)
                       More mismatches = slower but more comprehensive
                       4 mismatches: good balance of speed and sensitivity
        pam: PAM sequence (NGG for SpCas9)
    '''
    lines = [genome_path]

    # Build pattern: 20 N's for guide + PAM
    pattern = 'N' * 20 + pam
    lines.append(pattern)

    # Add each guide with mismatch tolerance
    for guide in guides:
        # Append NNN to represent PAM positions (not matched)
        lines.append(f'{guide}NNN {max_mismatches}')

    return '\n'.join(lines)