OpenMiChroM-Ana: Advanced Chromosome Structure Analysis Tool

Overview

OpenMiChroM-Ana is a powerful Python package designed for comprehensive analysis of chromosome structure data. It offers a suite of tools for processing, analyzing, and visualizing Hi-C and related genomic data. With support for both CPU and GPU acceleration, OpenMiChroM-Ana provides researchers with a flexible and efficient platform for exploring complex genomic structures.

Key Features

• Versatile Data Handling:

  • Support for Hi-C and simulated chromosome structure data
  • Efficient loading and preprocessing capabilities

• Comprehensive Analysis Tools:

  • Distance matrix calculations with multiple metrics
  • Advanced normalization methods (ICE, KR, VC, log transform)
  • State-of-the-art dimensionality reduction techniques (PCA, SVD, t-SNE, UMAP, MDS)
  • Diverse clustering algorithms (K-means, DBSCAN, Spectral, Hierarchical, OPTICS)
  • Robust clustering evaluation metrics

• Performance Optimization:

  • GPU acceleration for computationally intensive operations
  • Efficient CPU implementations for broad compatibility

• Visualization:

  • Rich set of plotting tools for result interpretation
  • Interactive visualizations for in-depth data exploration

Installation

System Requirements

• Python 3.11 or higher
• For GPU support: CUDA-Toolkit and CUDA version 12.0 or higher

CPU Version Installation

For users who prefer CPU-parallel operations:

# Navigate to the directory where the pyproject.toml is placed
pip install .

GPU Enabled Version

To leverage GPU acceleration:

1. Install CUDA Libraries (version 12.0 or higher)
2. Set up micromamba env or virtual env
3. Install RAPIDS Suite for CUDA ^12.0 (follow instructions at https://docs.rapids.ai/install)

mamba create -n [envName] -c rapidsai -c conda-forge -c nvidia rapids=24.06 python=3.11 cuda-version=12.0
mamba activate [envName]
# Navigate to the directory where the pyproject.toml is placed
pip install .[gpu]

Quick Start Guide

Initializing the Analysis

For CPU usage:

from OpenMiChroM_Ana import Ana

analysis = Ana(showPlots=True, execution_mode='cpu', cacheStoragePath='/path/to/cache')

For GPU usage:

from OpenMiChroM_Ana import Ana

analysis = Ana(showPlots=True, execution_mode='gpu', cacheStoragePath='/path/to/cache')

Basic Workflow Example

# Load datasets
analysis.add_dataset(label="ExperimentA", folder="data/ExperimentA")
analysis.add_dataset(label="ExperimentB", folder="data/ExperimentB")

# Process trajectory data
analysis.process_trajectories(label="ExperimentA", filename="traj_A.cndb", folder_pattern=['iteration_', [1, 20]])
analysis.process_trajectories(label="ExperimentB", filename="traj_B.cndb", folder_pattern=['iteration_', [1, 20]])

#NOTE: to cache the trajectories to avoid recomputation do 
# analysis.process_trajectories(label="ExpirementC" cache_trajs=True)

# Perform dimensionality reduction
pca_result = analysis.pca("ExperimentA", "ExperimentB", metric='euclidean', n_components=2, norm='ice', method='weighted')

# Conduct clustering analysis
kmeans_result = analysis.kmeans_clustering("ExperimentA", "ExperimentB", n_clusters=5, metric='euclidean', norm='ice', method='weighted')

# Visualize results
# Plots are automatically saved if showPlots=True

Contribution

We welcome contributions to OpenMiChroM-Ana! Whether it's bug fixes, feature additions, or documentation improvements, your input is valuable. Please review our contribution guidelines before submitting a pull request.

License

OpenMiChroM-Ana is distributed under the MIT License. See the LICENSE file in the repository for full details.

Support and Contact

For bug reports and feature requests, please use the GitHub issue tracker.