The article introduces a flexible R package for Nonnegative Matrix Factorization (NMF), designed to provide an open-source, user-friendly interface for standard NMF algorithms and a framework for implementing and testing new methods. NMF is an unsupervised learning technique widely applied in fields such as signal processing, face recognition, text mining, and bioinformatics. The package, developed for the R/BioConductor platform, includes several published NMF algorithms and initialization methods, facilitating the combination of these to create new NMF strategies. It also provides benchmark data and visualization tools for result comparison and interpretation. The package aims to address the limitations of existing commercial or proprietary implementations by offering a comprehensive and accessible solution for the bioinformatics community. The article details the implementation, features, and performance of the package, including its ability to handle large-scale data and its support for parallel computing. Examples and comparisons of different NMF methods are provided to demonstrate the package's capabilities.The article introduces a flexible R package for Nonnegative Matrix Factorization (NMF), designed to provide an open-source, user-friendly interface for standard NMF algorithms and a framework for implementing and testing new methods. NMF is an unsupervised learning technique widely applied in fields such as signal processing, face recognition, text mining, and bioinformatics. The package, developed for the R/BioConductor platform, includes several published NMF algorithms and initialization methods, facilitating the combination of these to create new NMF strategies. It also provides benchmark data and visualization tools for result comparison and interpretation. The package aims to address the limitations of existing commercial or proprietary implementations by offering a comprehensive and accessible solution for the bioinformatics community. The article details the implementation, features, and performance of the package, including its ability to handle large-scale data and its support for parallel computing. Examples and comparisons of different NMF methods are provided to demonstrate the package's capabilities.