Final Error Correction Benchmarking Repository

Description of project directories

• code.evaluation:

  • data_analysis:
    • data: contains the final, compressed data used for analysis.
    • IGH_Analysis.ipynb, RSR_Analysis.ipynb, T1_Analysis.ipynb: Jupyter nobebooks used to generate figures and analysis of the final compressed data generated from /code.evaluation/evaluation/ec_summary.py which was called after each tool was ran in the code.evaluation/job_submission/master_wrapper.sh
  • job_submission: contains information on how to submit jobs for all/sets of the tools on the hoffman cluster for each of the datasets evaluated in this study.
    • see README.md in this directory.
  • evaluation:
    • ../job_submission/master_wrapper.sh calls ec_evaluation.py which calls ec_data_compression.py
    • ../job_submission/master_wrapper.sh then calls ec_summary.py with the files created in ec_data_compression.py
    • data_metrics.py generates metrics such as precision, gain, accuracy, and sensitivity from the summary data.
  • testing_data:
    • shows an example of small file set to be ran
    • displays examples of compressed file output from evaluation.
  • wrappers:
    • .sh wrappers used to run all tools, produce standard log files.
    • contains the wrappers for the tools used.
    • wrappers convert the reads to a merged input file.
    • single end reads are ok for paired end wrappers by passing blank.fastq.

• gold.standard:

  • rep_seq_real:
    • contains the scripts for submitting the jobs to create the UMI based gold standard reads for the TCR data. (generate_files.sh, rep_seq_real_to_true.py)
    • contains the scripts for visualizing the UMI grouping sizes. (rep_seq_visualizer.py)
    • contains the scripts for retrieving SRA data using qsubs on the hoffman cluster (handle_dump.sh)

• true.reads: