Input source for mosquito short read alignment pipeline

[alimanfoo]

For the mosquito short read alignment pipeline, I believe there are two main options regarding the input data source for the sequence reads:

• FASTQ files from ENA
• BAM/CRAM files from IRODS

There are some practical trade-off in these choices:

• Using FASTQ as input is simpler because no prior conversion is required; using BAM/CRAM makes for a bit more pipeline engineering, because the BAM/CRAM needs to be converted to FASTQ
• Using FASTQ from ENA is more portable as ENA is accessible from anywhere whereas IRODS is only accessible from within WSI network (I believe)
• Using IRODS may be faster because data transfer to WSI lustre may be faster, either because of better network connectivity or more efficient transfer protocol -- this needs to be investigated and confirmed

[alimanfoo]

cc @kemprg @gbggrant @tnguyensanger @JonKeatley112 @hardingnj

[gbggrant]

If we use the bam/crams and convert them into fastqs for input to the short read alignment pipeline, we lose all of the tags at that step. Do we then want to add a MergeBamAlignment step after bwa to add those tags back to the aligned bam?

[alimanfoo]

Assuming this is resolved and can be closed.