Afann

Afann (Alignment-Free methods Adjusted by Neural Network) is a software tool that improves the accuracy of alignment-free methods for comparing genome sequences or raw sequencing samples without assembly. Alignment-free methods are faster and more memory-efficient than alignment-based methods. Still, they can overestimate the dissimilarity between sequencing samples and the dissimilarity calculated based on their genomes. This bias can significantly decrease the performance of the alignment-free analysis. Afann addresses this issue by using a neural network to adjust the bias, improving performance on various independent datasets.

Topic

Machine learning;Sequencing;Sequence assembly

Detail

  • Operation: Standardisation and normalisation;k-mer counting;Regression analysis

  • Software interface: Command-line user interface

  • Language: C++,Python

  • License: Other

  • Cost: Free fo available for academic or non-commercial purposes only.

  • Version name: v1.0.0

  • Credit: US National Science Foundation, National Institutes of Health.

  • Input: -

  • Output: -

  • Contact: Fengzhu Sun fsun@usc.edu

  • Collection: -

  • Maturity: -

Publications

  • Afann: bias adjustment for alignment-free sequence comparison based on sequencing data using neural network regression.
  • Tang K, et al. Afann: bias adjustment for alignment-free sequence comparison based on sequencing data using neural network regression. Afann: bias adjustment for alignment-free sequence comparison based on sequencing data using neural network regression. 2019; 20:266. doi: 10.1186/s13059-019-1872-3
  • https://doi.org/10.1186/S13059-019-1872-3
  • PMID: 31801606
  • PMC: PMC6891986

Download and documentation


< Back to DB search