Version Compatibility

Reference examples tested with: numpy 1.26+, pandas 2.2+, pysam 0.22+

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

• Python: pip show <package> then help(module.function) to check signatures

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

Fragment Analysis

"Analyze cfDNA fragment patterns for cancer detection" → Extract fragmentomics features (size distributions, nucleosome positioning, DELFI profiles) from cfDNA for tumor detection and tissue-of-origin analysis.

• Python: FinaleToolkit or Griffin for fragment feature extraction
• Python: pysam for custom fragmentomics analysis

Analyze cfDNA fragmentomics for cancer detection and characterization.

Tool Selection

Tool	Description	Use Case
FinaleToolkit	DELFI-style patterns, MIT license	General fragmentomics
Griffin	Nucleosome profiling	Tissue deconvolution

Note: DELFI is a commercial company, NOT software. Use FinaleToolkit (MIT license) which replicates DELFI patterns and is 50x faster.

Fragment Size Metrics

import pysam
import numpy as np
import pandas as pd


def calculate_fragment_metrics(bam_path):
    '''
    Calculate cfDNA fragment metrics.

    Key ratios for cancer detection:
    - Short (100-150 bp) vs Long (151-220 bp)
    - ctDNA tends to be shorter than normal cfDNA
    '''
    bam = pysam.AlignmentFile(bam_path, 'rb')
    sizes = []

    for read in bam.fetch():
        if read.is_proper_pair and not read.is_secondary and read.template_length > 0:
            sizes.append(read.template_length)

    bam.close()
    sizes = np.array(sizes)

    # DELFI-style ratios
    short = np.sum((sizes >= 100) & (sizes <= 150))
    long = np.sum((sizes >= 151) & (sizes <= 220))

    metrics = {
        'total_fragments': len(sizes),
        'median_size': np.median(sizes),
        'mean_size': np.mean(sizes),
        'short_fragments': short,
        'long_fragments': long,
        'short_long_ratio': short / long if long > 0 else np.nan,
        # Mononucleosome peak
        'mono_peak_fraction': np.sum((sizes >= 150) & (sizes <= 180)) / len(sizes)
    }

    return metrics