Skip to content

Latest commit

 

History

History
executable file
·
88 lines (77 loc) · 4.67 KB

README.md

File metadata and controls

executable file
·
88 lines (77 loc) · 4.67 KB

Usage Example

This page shows an example of applying the procedure described in Episomizer page to construct circular DNA structures on the mini-bam file created by simply keeping the reads around the CNA segment boundaries of the relapse sample from a pediatric HGG patient (SJHGG019_S). For the general usage information, see Episomizer home page.

Set Environment

Set $EPISOMIZER_HOME to the cloned directory, add bin path to $PATH and enter $EPISOMIZER_HOME directory.

$ EPISOMIZER_HOME=<path_to_cloned_dir>
$ export PATH=$EPISOMIZER_HOME/bin:$PATH
$ cd $EPISOMIZER_HOME

Procedure

Step 1: We used Log2Ratio > 4 based on the CONSERTING Log2Ratio distribution to determine the highly amplified genomic segments (file CNA_region_raw_R.bed). For descriptions about all the example input files, see input page.

Step 2: Get the putative edges.

  1. Generate the shell script with samtools commands and the intermediate folder (CNA_boundary_reads) in the given output directory.

    $ episomizer create_samtools_cmd ./testdata/input/mini_SJHGG019_S.bam ./testdata/input/CNA_region_raw_R.bed ./testdata/intmd/trace

    Run the shell script.

    $ ./testdata/intmd/trace/run_samtools.sh

  2. Generate the shell script to extract softclip reads and the intermediate folder (CNA_boundary_softclip_fa) in the given output directory.

    $ episomizer create_softclip2fa_cmd ./testdata/input/CNA_region_raw_R.bed ./testdata/intmd/trace

    Run the shell script.

    $ ./testdata/intmd/trace/run_softclip2fa.sh

  3. Generate the shell script to blat the softclip reads and the intermediate folder (CNA_boundary_softclip_BLAT) in the given output directory.

    $ episomizer create_blat_cmd GRCh37-lite.2bit ./testdata/input/CNA_region_raw_R.bed ./testdata/intmd/trace

    The reference genome GRCh37-lite.2bit can be downloaded from St. Jude public FTP site and can be placed under the working directory.

    Run the shell script (submitting the jobs in parallel is strongly recommended).

    $ ./testdata/intmd/trace/run_BLAT.sh

  4. Create 3 read count matrices using softclip reads, discordant reads and bridging discordant reads.

    $ episomizer SV_softclip ./testdata/input/CNA_region_raw_R.bed 1000 ./testdata/intmd/trace/CNA_boundary_softclip_BLAT ./testdata/intmd/trace
$ episomizer SV_discordant ./testdata/input/CNA_region_raw_R.bed 800 1000 ./testdata/intmd/trace/CNA_boundary_reads ./testdata/intmd/trace
$ episomizer SV_bridge ./testdata/input/CNA_region_raw_R.bed 800 10000 ./testdata/intmd/trace/CNA_boundary_reads ./testdata/intmd/trace

  5. Convert matrix file to edges file.

    $ episomizer matrix2edges ./testdata/input/CNA_region_raw_R.bed ./testdata/intmd/trace/matrix_softclip.txt ./testdata/intmd/trace/putative_edges_softclip.txt
$ episomizer matrix2edges ./testdata/input/CNA_region_raw_R.bed ./testdata/intmd/trace/matrix_discordant.txt ./testdata/intmd/trace/putative_edges_discordant.txt
$ episomizer matrix2edges ./testdata/input/CNA_region_raw_R.bed ./testdata/intmd/trace/matrix_bridge.txt ./testdata/intmd/trace/putative_edges_bridge.txt

Step 3: Manually review the putative edges. Please follow the instruction in Episomizer home page. The review process for the relapse sample is summarized in edges_review_relapse_rawCNA.xlsx.

Step 4: Compose circular double minute structures.

$ episomizer composer circ -c ./testdata/intmd/reviewed_data/Reviewed_segments_for_graph_relapse_rawCNA.bed -l ./testdata/intmd/reviewed_data/Reviewed_edges_for_graph_relapse_rawCNA.txt -d ./testdata/output/double_minutes_relapse.txt

The constructed double minutes are in double_minutes_relapse.txt.

Notes on the diagnosis sample (SJHGG019_E)

We used the same workflow to generate putative edges for the diagnosis sample, but since the sequencing coverage for the diagnosis sample is much choppier, the CNA segments boundaries are not as accurate as the relapse sample, which influenced the identification of a few SVs. After manual review of the reads and their Blat results, together with the Chromium data, we rescued eight missed SVs. The review process can be viewed at edges_review_diagnosis_rawCNA.xlsx.