Tool for pipeing inputs and outputs of multiple cli tools. Pigeon takes in only a config file as input. Everything required to run the pipeline are specified in config file. The config file is specified according to python configparser.

sudo pip3 install --index-url https://test.pypi.org/simple/ pigeon

pip install --index-url https://test.pypi.org/simple/ pigeon

None of the tools or data files are supplemented by pigeon so they need to be downloaded. For example configuration file, exome sequencing pipeline,

• Tools
  • FastQC
  • BWA
  • Picard
  • GATK
  • Annotation Tool
• Reference Files
  • Reference genome and known SNP&INDELS
    • hg19 or
    • hg38
  • Bed file
    • See website of capture kit used in sequencing

Create yourself a configuration file

pigeon createconfig

Modify for your analysis. (See below.)

pigeon -c my_config.conf -d

If everything looks alright run for real.

pigeon -c my_config.conf

Config file consists of three parts.

• General
• Pipeline
• Individual tool blocks

Area used to define project name, output directory, input files, and resource files like reference genome or target file. Following variables are necessary for run.

Required:

• project_name : name of your project

• output_dir : where to write output files

• input_files : input files for analysis, space separated, pairs should be next to each other. e.g.

  input_files = A.txt B.txt C.txt

  or for paired

  input_files = A1.txt A2.txt B1.txt B2.txt C1.txt C2.txt

Optional variable can also be decleared here. Based on your or tool requirements. Later these variables can be called in the config file using ${GENERAL:optional_variable}.

Optional(example):

reference_genome = /path/to/my/reference_genome.fa

bed_file = /path/to/my/target.bed

known_snp = /path/to/my/snp.vcf

my_database = /path/to/my/favorite.db

This area should contain paths to tools that is understanble by your shell. As well as the run order of tools. e.g.

pipeline = job1 job2 job3

A = path/to/A

B = path/to/B

C = path/to/C

Name of the block should be same as in pipeline. By continuing example above;

[job1]

[job2]

[job3]

Arguments that can be used in these blocks as follows:

• tool: tool variable from pipeline block. e.g.

  tool = ${PIPELINE:A}

• sub_tool: if tool has a sub tool like 'bwa mem'. e.g.

  sub_tool = mem

• args: arguments of the tools

• java: if tool is a jar file add java -jar before it.

• pass: if True it won't run the block. But the block still be part of the pipeline. This option is helpful for resuming interrupted pipeline.

• input_from: Name of the block that that's output is this jobs input. First jobs input_from should be input_files.

• input_multi: can be 'paired' or 'all'. Paired option splits input files stream into groups of two. All option uses all of the input files.

• input_flag_repeat: If tool requires input flag for each input this command will add given flag before each input.

• secondary_in_placeholder

• suffix: add suffix to output file name

• ext: file extension of the output

• dump_dir: creates a directory and outputs there.

• paired_output: this option will pair the input and the output of the tool.

• secondary_out_placeholder

• secondary_suffix

• secondary_ext

• secondary_dump_dir

These are joker words that can be used in args.

• input_placeholder

• secondary_input_placeholder

• output_placeholder

• secondary_output_placeholder

[[GENERAL]]
project_name = my_project
output_dir = /path/to/output_directory
input_files = A.txt B.txt C.txt
my_db = /path/to/my.db

[[PIPELINE]]
pipeline = job1 job2 job3
A = /path/to/A
B = /path/to/B
C = /path/to/C

[job1]
tool = A
input_from = input_files
args = -i input_placeholder -o output_placeholder
suffix = job1_A
ext = txt

[job2]
tool = B
input_from = job1
args = -i input_placeholder -o output_placeholder
suffix = job2_B
ext = txt

[job3]
tool = C
input_from = job2
args = -i input_placeholder -o output_placeholder
suffix = job3_C
ext = txt