WIP: production hotfixes #98
Conversation
| # if successful, set self.reports_path | ||
| self.reports_path = join(self.pipeline.output_path, | ||
| 'SeqCountsJob', | ||
| 'SeqCounts.csv') |
There was a problem hiding this comment.
Is this file consistent in schema with Demultiplex_Stats.csv used in the other object for reports_path?
There was a problem hiding this comment.
No Demultiplex_Stats.csv's schema is defined and maintained by Illumina and carries additional stats beyond sequence counts. SeqCounts.csv is defined and maintained by us and just records the total raw-read counts for forward and reverse and the lane number per sample.
Consistency in this case is not an issue since the consumer of the metadata is the metapool module, and it only uses Demultuplex_Stats.csv for its raw-read counts. The equivalent output file for the bcl2fastq is similarly different. Our new method is more accurate in that it counts both the forward and reverse reads while I believe in Demultuplex_Stats.csv we just double the forward count.
There was a problem hiding this comment.
Isn't it risky, and unexpected, for an attribute common across subclasses to have a different interpretation?
| # shared method, but probably still in Assay. | ||
| self.generate_prep_file() | ||
|
|
||
| self.generate_prep_file() |
There was a problem hiding this comment.
It seems like the stuff in this method is the same as what's in the StandardMetagenomicWorkflow, can execute_pipeline be pushed to a base class?
There was a problem hiding this comment.
I could push much of both into the MetaOmics base class off the top of my head. Those two child classes are very similar. Currently they differ only in the string constants we define, off the top of my head. More than happy to include that in my next push.
Production hotfixes. Tests not yet updated.