Could you briefly describe how different run_acces...
# taxprofiler
m
Could you briefly describe how different run_accessions are treated in the pipeline, I ran
test
and I'm confused. the
analysis_ready_fastqs
directory shows merged and umapped. In the samplesheet for
sample 2612
we have
run_accession ERR5766176
which is pair end. But I cannot understand if the pair ends are also merged with diff run_accessions?
Copy code
$ tree analysis_ready_fastqs/
analysis_ready_fastqs/
├── 2612_ERR5766176_B.unmapped.fastq.gz
├── 2612_ERR5766176.unmapped.fastq.gz
├── 2612_ERR5766180.unmapped.fastq.gz
├── 2612.merged.fastq.gz
└── 2613_ERR5766181.unmapped.fastq.gz
Copy code
csv
sample,run_accession,instrument_platform,fastq_1,fastq_2,fasta
2611,ERR5766174,ILLUMINA,,,<https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fasta/ERX5474930_ERR5766174_1.fa.gz>
2612,ERR5766176,ILLUMINA,<https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474932_ERR5766176_1.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474932_ERR5766176_2.fastq.gz>,
2612,ERR5766176_B,ILLUMINA,<https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474932_ERR5766176_B_1.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/>
data/fastq/ERX5474932_ERR5766176_B_2.fastq.gz,
2612,ERR5766180,ILLUMINA,<https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474936_ERR5766180_1.fastq.gz>,,
2613,ERR5766181,ILLUMINA,<https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERX5474937_ERR5766181_1.fastq.gz,https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data>
/fastq/ERX5474937_ERR5766181_2.fastq.gz,
ERR3201952,ERR3201952,OXFORD_NANOPORE,<https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fastq/ERR3201952.fastq.gz>,,
2614,ERR3201952,OXFORD_NANOPORE,,,<https://raw.githubusercontent.com/nf-core/test-datasets/taxprofiler/data/fasta/ERR3201952_dup.fa.gz>
j
I will get back tomorrow @m3hdad! Some details I need to check first as I don't remember and with my kids now
👍 1
OK still not had much time to go through the logic really closely, but conceptually: 1. Reads are merged if paired end (and pair-end collapsing turned on) per run-accession separarely 2. Host reads are then removed for merged-pairs and single end per run accession separately 3. All files that have the same sample is are then merged I think (I need to check but kiddo won't let me) that if you don't merge pairs, then single-end and paired-end will only be merged on a like basis (i.e. single-ends will be run merged togtehr, paired-end will be run merged together, but the resulting 'merged' files will not be merged together - the single-end merged and paired-end merged will remain as separate files for downstream
Does that help?
m
Yes, thanks James. Understood the logic.
👍 1
j
Does that system work for you?
m
Yes, I believe the pipeline is flexible enough to let the user decide. Now I see that there are two parameters which control this behavior
--perform_runmerging
and
--shortread_qc_mergepairs
.
👍 1