Feature #4042
Updated by Tom Clegg over 9 years ago
Example job patterns that should be supported: * 6 directories, each with 8 pairs of fastq files → 48 tasks, each operating on one pair * 3 dirs with 1 pair of fastq files each + 3 dirs with 4 pairs of fastq files each → 15 tasks, each operating on one pair * 6 dirs, each with 8 bam files → 6 tasks, each operating on one dir * 6 dirs, each with 1 bam file, plus dirs + a static list of chromosomes → 6×23 23×6 tasks, each operating on one {bam file, {dir, chromosome} * 6 dirs, each with 23 bam files → 6×23 tasks, each operating on one file * 6 dirs, each with 23 bam files → 6 tasks, each operating on (all of the bam files in) one dir The output should preserve the input directory structure: in each of the above examples, the output collection should have 6 directories with the same names as the 6 input directories.