BOAssembler

BOAssembler automates hyper-parameter tuning of RNA-Seq assemblers using Bayesian Optimization to improve recovery of transcript sequences from short fragments produced by high-throughput sequencing.


Key Features:

  • Bayesian Optimization: Uses Bayesian Optimization principles to guide selection of assembler hyper-parameters.
  • Automatic hyper-parameter tuning: Performs end-to-end automatic tuning of RNA-Seq assembler parameters.
  • Transcript reconstruction focus: Targets recovery of original RNA transcripts from millions of short fragments generated by high-throughput sequencing.
  • Improved assembly performance: Has been demonstrated experimentally to improve assembly performance across multiple datasets.
  • Facilitates downstream analysis: Produces optimized assemblies that support more accurate gene, protein, and cell analyses.
  • Applicability to benchmarking: Provides a data-driven approach applicable to bioinformatics benchmark studies.

Scientific Applications:

  • Transcriptome assembly optimization: Optimizes assembler parameters to improve transcriptome assembly quality from RNA-Seq data.
  • Dataset-specific tuning: Enables dataset-specific parameter optimization to achieve consistent performance across diverse RNA-Seq datasets.
  • Downstream molecular analysis: Enhances the accuracy of downstream gene, protein, and cell-level analyses by improving assembly quality.
  • Bioinformatics benchmarking: Serves as a method for systematic evaluation and comparison of assembler performance in benchmark studies.

Methodology:

Applies Bayesian Optimization to tune hyper-parameters of RNA-Seq assemblers for transcript reconstruction from short fragments generated by high-throughput sequencing.

Topics

Details

Tool Type:
command-line tool
Programming Languages:
Python
Added:
1/18/2021
Last Updated:
11/24/2024

Operations

Publications

[No authors listed]. BOAssembler: A Bayesian Optimization Framework to Improve RNA-Seq Assembly Performance. Algorithms for Computational Biology. 2020;12099:188.

PMCID: PMC7197064