Skip to content

Evaluation of the statistical reproducibility of high-throughput structural analyses for a robust RNA reactivity classification

License

Notifications You must be signed in to change notification settings

carushi/reactIDR

Repository files navigation

reactIDR: evaluation of the statistical reproducibility of high-throughput structural analyses towards a robust RNA structure prediction

  • Input read count data

    • PARS
    • SHAPE-Seq
    • icSHAPE
    • DMS-Seq (assumed to be enriched only at A or C)
  • Output

    • posterior probability of being loop (enriched in case) or stem (enriched in control)
  • Algorithm

    • IDR + hidden Markov Model

Requirement

  • python3
  • numpy
  • scikit-learn

Other packages are required for visualization process as follows:

  • pandas
  • seaborn

How to start

git clone https://github.com/carushi/reactIDR
cd reactIDR/
python setup.py  build_ext -b reactIDR/ # cython build
cd example && bash training.sh    # Run test

Please visit our wiki for further info.

Script

  • read_collapse.py
    • collapse PCR duplicates and trim barcode
    • assume gawk
  • read_truncate.py
    • extract consistent paired end reads
  • bed_to_pars_format.py
    • write PARS-formatted 5' end coverage data based on gtf and gff annotation or sequence location
    • format: NAME 0;1;2;3;.....
  • tab_to_csv.py
    • use to append raw count (read count, coverage, ...) to the output csv file

Reference

TODO

  • apply to MaP analyses

About

Evaluation of the statistical reproducibility of high-throughput structural analyses for a robust RNA reactivity classification

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published