telomere-finder: a tool to identify putative telomeric and subtelomeric sequences from draft genome assemblies

Tutorial

The tool telomere-finder.py helps identify potential telomeric repeats in genome assemblies by analyzing k-mer occurrences at contig ends. It supports both compressed and uncompressed FASTA inputs and can output results in BED or GFF format.

Installation

# Clone the repository
git clone https://github.com/yourusername/telomere-finder
cd telomere-finder

# Install dependencies
pip install biopython

Usage

Basic usage with default settings (6-mers and 7-mers, BED output):

python telomere_finder.py input.fasta

Analyze specific k-mer sizes and output GFF:

python telomere_finder.py --kmers 6 7 8 --format gff input.fasta

Demo Example

Let's analyze a simple demo genome with known telomeric repeats. The demo file demo.fasta contains three chromosomes:

• chr1: Contains multiple TTTAGGG repeats at both ends
• chr2: Has TTAGGG repeats internally and CCCTAA at the end
• chr3_reverse: Contains the canonical telomere sequence in reverse orientation (CCCTAA)

# Run analysis on demo genome
python telomere_finder.py --kmers 6 7 demo.fasta

Expected output:

Top telomeric repeats found at contig ends:
TTAGGG: 12 occurrences at contig ends
TTTAGGG: 8 occurrences at contig ends
CCCTAA: 12 occurrences at contig ends

Output Files

1. .bed or .gff file containing positions of all k-mer occurrences:

   • BED format (0-based): chromosome start end kmer score strand
   • GFF format (1-based): Includes additional metadata about end proximity TODO double-check basing. FOLLOWUP-20250101: In "telomere-finder-2.py" I corrected line 63.

2. .stats.txt file containing:

   • K-mer occurrence counts
   • Number of times each k-mer appears near contig ends
   • Distribution of k-mers across the genome

Finding Telomeres

The tool is particularly useful for:

1. Identifying potential telomeric repeats in draft assemblies
2. Validating genome assembly completeness
3. Detecting non-canonical telomere sequences
4. Analyzing telomere orientation and distribution

Note that the tool considers sequences within 1000bp of contig ends as potential telomeric regions by default.

1. First, this variable in the constructor defines what counts as an "end":

def __init__(self, kmer_sizes, end_region_size=1000):
    self.end_region_size = end_region_size

2. Then in the process_sequence method, a k-mer is marked as being at an "end" if it meets this condition:

'is_end': i < self.end_region_size or i > seq_len - self.end_region_size

Breaking this down:

• i < self.end_region_size: This checks if the k-mer starts within the first 1000bp of the sequence
• i > seq_len - self.end_region_size: This checks if the k-mer starts within the last 1000bp of the sequence

Any k-mer that occurs within 1000 bases of either the start or end of a contig is considered to be at an "end". This is somewhat arbitrary - telomeres can be longer or shorter than 1000bp.

Advanced Usage

Compressed Input Files

The tool automatically handles gzipped FASTA files:

python telomere_finder.py input.fasta.gz

Custom End Region Size

Modify the end_region_size parameter in the code to adjust what's considered an "end region" (default: 1000bp).

Output Format Conventions

• BED format: 0-based, half-open intervals
• GFF format: 1-based, closed intervals

Common Telomeric Sequences

Human and many vertebrates:

• TTAGGG (canonical)
• CCCTAA (reverse complement)

Some known variants:

• TTTAGGG
• TTAGGGG
• TTAGG

Tips for Analysis

1. Start with multiple k-mer sizes to catch variants
2. Look for reverse complement sequences
3. Pay attention to k-mer clustering at contig ends
4. Consider both frequency and position of repeats
5. The current output (as of December 10, 2024) identifies top kmers in the CLI summary, which might not match canonical telomeric repeats like TTAGGG.