Automated nf-core/eager processing of Autorun output bams.
-
Run
prepare_eager_tsv.Rfor human SG or TF data for a given sequencing batch:prepare_eager_tsv.R -s <batch_Id> -a SG -o eager_inputs/ -d .eva_credentials prepare_eager_tsv.R -s <batch_Id> -a TF -o eager_inputs/ -d .eva_credentials
-
Run eager with the following script, which then runs on the generated TSV files:
run_Eager.sh
Contains the autorun, SG and TF profiles.
Broader scope options and parameters for use across all processing with autorun.
Turns off automatic cleanup of intermediate files on successful completion of a run to allow resuming of the run when additional data becomes available, without rerunning completed steps.
This is profile contains all paramters provided to BOTH SG and TF runs that are paths to files on the local MPI-EVA filesystem.
They are provided in a separate profile to make it clearer to provided added transparency as well as make it easier for third parties to reproduce the processing done with Autorun_eager (e.g. by loading the SG or TF remote profiles) without getting errors about paths that do not exist on their filesystem.
The standardised parameters for processing human shotgun data.
The standardised parameters for processing human 1240k capture data.
An R script that when given a sequencing batch ID, Autorun Analysis type and PANDORA credentials will create/update eager input TSV files for further processing.
Usage: ./prepare_eager_tsv.R [options] .credentials
Options:
-h, --help
Show this help message and exit
-s SEQUENCING_BATCH_ID, --sequencing_batch_id=SEQUENCING_BATCH_ID
The Pandora sequencing batch ID to update eager input for. A TSV file will be prepared
for each individual in this run, containing all relevant processed BAM files
from the individual
-a ANALYSIS_TYPE, --analysis_type=ANALYSIS_TYPE
The analysis type to compile the data from. Should be one of: 'SG', 'TF'.
-r, --rename
Changes all dots (.) in the Library_ID field of the output to underscores (_).
Some tools used in nf-core/eager will strip everything after the first dot (.)
from the name of the input file, which can cause naming conflicts in rare cases.
-w WHITELIST, --whitelist=WHITELIST
An optional file that includes the IDs of whitelisted individuals,
one per line. Only the TSVs for these individuals will be updated.
-o OUTDIR/, --outDir=OUTDIR/
The desired output directory. Within this directory, one subdirectory will be
created per analysis type, within that one subdirectory per individual ID,
and one TSV within each of these directory.
-d, --debug_output
When provided, the entire result table for the run will be saved as '<seq_batch_ID>.results.txt'.
Helpful to check all the output data in one place.
Note: a valid sidora .credentials file is required. Contact the Pandora/Sidora team for details.The eager input TSVs will be created in the following directory structure, given -o eager_inputs:
eager_inputs
├── SG
│ └──IND
│ ├── IND001
│ └── IND002
└── TF
└──IND
├── IND001
└── IND002
Alongside each created TSV is a file named autorun_eager_version.txt, which states the version of Autorun_eager used.
A wrapper shell script that goes through all TSVs in the eager_inputs directory, checks if a completed run exists for a given TSV, and submits/resumes an
eager run for that individual if necessary.
Currently uses eager version 2.4.5 and profiles eva,archgen,medium_data,autorun,local_paths across all runs, with the SG or TF profiles used for their respective
data types.
The outputs are saved with the same directory structure as the inputs, but in a separate parent directory.
eager_outputs
├── SG
│ └──IND
│ ├── IND001
│ └── IND002
└── TF
└──IND
├── IND001
└── IND002
This script recognises the -a/--array option. When this is provided, instead of running eager jobs in sequence, a temporary file is created named $(date +'%y%m%d_%H%M')_Autorun_eager_queue.txt that includes the command line of all eager jobs to-be-ran, one per line. An "Autorun_eager spawner" (AE_spawner) array job is then submitted using qsub, which uses a secondary script named scripts/submit_as_array.sh to submit the command in each line of the temporary file as a separate task. In this manner, 10 eager runs can be ran in parallel. Logs for these jobs will then be added to a directory named array_Logs/<temp_file_name>/.
A shell script that will check the available TF genotype datasets for a given individual and create/update a poseidon package with the data of that individual.
usage: ./scripts/update_poseidon_package.sh [options] <ind_id>
This script pulls data and metadata from Autorun_eager for the TF version of the specified individual and creates a poseidon package.
Options:
-h, --help Print this text and exit.
-v, --version Print version and exit.
Comparing the timestamp of the Autorun_eager genotypes and those in the poseidon package for the individual (if one exists already) this script will create/update a poseidon package where necessary by doing the following:
- Check if a package creation/update is needed. If not, exit happily.
- Create a bare poseidon package from the available Autorun_eager genotypes in a temporary directory.
- Paste together dsDNA and ssDNA genotypes before creating bare package, if necessary.
- Pull information from Pandora and Autorun_eager to populate the package janno files, using
scripts/fill_in_janno.R. This also updates the Genetic_Sex field. - Update the genetic sex in the package ind file, so it matches the updated janno, using
scripts/update_dataset_from_janno.R. - Use
trident updateto bump the package version (1.0.0if the package is newly created), and create a Changelog. - Validate the resulting package.
- If validation passes, publish the (updated) version of the package to the central repository in
poseidon_packages/and remove any temporary files created.
A shell script that scrubs the Autorun_eager input and output directories of all individuals in a specified list of sensitive sequencing IDs. This is used daily with the most up-to-date list of sensitive sequencing IDs to ensure that no results are available even if marking samples as sensitive was done late.
usage: ethical_sample_scrub.sh [options] <sensitive_seqIds_list>
This script pulls the Pandora individual IDs from the list of sensitive sequencing IDs, and
removes all Autorun_eager input and outputs from those individuals (if any).
This ensures that no results are available even if marking samples as sensitive was done late.
Options:
-h, --help Print this text and exit.
A shell script that will clear the work directories of individuals in a specified individual ID list from both the SG and TF results directories.
usage: clear_work_dirs.sh [options] <ind_id_list>
This script clears the work directories of individuals in a specified individual ID list from both the SG and TF results directories.
Options:
-h, --help Print this text and exit.
A shell script that clears the results directories of all individuals in a specified list While maintaining nextflow's caching of already-ran processes. This is useful for refreshing the results directories of individuals when changes to the input might have changes merging of libraries, thus making the directory structure inconsistent.
usage: clear_results.sh [options] <ind_id_list>
This script removes all output directory contents for the provided individuals, without clearing out caching, allowing for the results to be re-published.
This enables refreshing of result directories when changes to the input might have changes merging of libraries, thus making the directory structure inconsistent.
Options:
-h, --help Print this text and exit.
-a, --analysis_type Set the analysis type. Options: TF, SG.