Welcome to the CNV-Benchmarking repository. This project is dedicated to comparing and evaluating some of the leading algorithms in the field of Copy Number Variation (CNV) analysis: CONICSmat, HoneyBADGER, and InferCNV. Our goal is to provide a comprehensive benchmarking analysis that helps researchers and practitioners in selecting the most suitable tool for their CNV studies.
Copy Number Variations (CNVs) are significant in genomic studies and have implications in various genetic disorders and cancers. Accurate detection and analysis of CNVs are crucial. This repository focuses on benchmarking three prominent CNV analysis algorithms:
- CONICSmat: Designed for detecting CNVs in single-cell RNA sequencing data.
- HoneyBADGER: Known for its ability to identify CNVs and allele-specific expression concurrently.
- InferCNV: A tool that offers insights into CNV events from scRNA-seq data.
The benchmarking is conducted based on several key criteria:
- Accuracy: Evaluating the true positive and false positive rates.
- Scalability: Assessing the performance with varying data sizes.
- Usability: Considering the ease of use and documentation.
- Resource Efficiency: Analyzing the computational resource requirements.
(Here, you can briefly summarize the results or key findings of the benchmarking process. Alternatively, point users to a specific section or document where detailed results are available.)
To replicate the benchmarking study or to use these tools for your own analysis, follow these steps:
- Clone the Repository:
git clone [repository URL] - Install Dependencies: (List out any necessary dependencies and installation instructions)
- Run Benchmarking Tests: (Provide instructions on how to execute the benchmarking tests)
We welcome contributions from the community! If you have suggestions, improvements, or additional tools to benchmark, please feel free to fork this repository, make your changes, and submit a pull request.
This project is licensed under [specify the license type here].
For any inquiries or support, please contact [email protected] or raise an issue in the repository.