matchms

name	matchms
description	Find the most similar spectral pair from MSP files using CosineGreedy similarity. Use when comparing multiple mass spectra, computing pairwise similarity matrices, or identifying best matches in metabolomics.

Matchms: Find Most Similar Spectral Pair

Overview

Given an MSP file with multiple spectra, find the pair with the highest CosineGreedy similarity score.

Loading MSP Files with Multiple Spectra

from matchms.importing import load_from_msp

def load_all_spectra(filepath):
    """Load all spectra from an MSP file."""
    return list(load_from_msp(filepath))

Computing Pairwise Similarities

from matchms.similarity import CosineGreedy
from matchms.importing import load_from_msp

# Load all spectra
spectra = load_all_spectra("/root/input.msp")

# Compute pairwise similarities
cosine_greedy = CosineGreedy(tolerance=0.1)

best_pair = None
best_score = -1

for i in range(len(spectra)):
    for j in range(i + 1, len(spectra)):
        score = cosine_greedy.pair(spectra[i], spectra[j])
        if score > best_score:
            best_score = score
            best_pair = (spectra[i].get("compound_name", spectra[i].get("name", f"Spectrum_{i}")),
                        spectra[j].get("compound_name", spectra[j].get("name", f"Spectrum_{j}")))

# Write result
with open("/root/output.txt", "w") as f:
    f.write(f"Most similar pair: {best_pair[0]} and {best_pair[1]}\n")
    f.write(f"Similarity: {best_score:.3f}\n")

Complete Script

from matchms.importing import load_from_msp
from matchms.similarity import CosineGreedy

# Load all spectra from MSP file
spectra = list(load_from_msp("/root/input.msp"))

# Initialize similarity calculator
cosine_greedy = CosineGreedy(tolerance=0.1)

# Find most similar pair
best_score = -1
best_pair = (None, None)

for i in range(len(spectra)):
    for j in range(i + 1, len(spectra)):
        score = cosine_greedy.pair(spectra[i], spectra[j])
        if score > best_score:
            best_score = score
            name_i = spectra[i].get("compound_name", spectra[i].get("name", f"Spectrum_{i}"))
            name_j = spectra[j].get("compound_name", spectra[j].get("name", f"Spectrum_{j}"))
            best_pair = (name_i, name_j)

# Write result
with open("/root/output.txt", "w") as f:
    f.write(f"Most similar pair: {best_pair[0]} and {best_pair[1]}\n")
    f.write(f"Similarity: {best_score:.3f}\n")

Output Format

Most similar pair: {name1} and {name2}
Similarity: X.XXX

Key Reference

• load_from_msp(filepath) — Load all spectra from MSP file
• CosineGreedy(tolerance=0.1) — Create similarity calculator (tolerance in m/z)
• cosine_greedy.pair(spec1, spec2) — Compute similarity between two spectra
• spectrum.get("compound_name") — Get spectrum name from MSP Name: field (matchms stores it as 'compound_name')