This is the wrong behavior, because those files have the rows FIRST_OF_PAIR and SECOND_OF_PAIR as well, which should have also been added to the JSON. This is because the picardTools parser uses self.library_read_type to determine whether reads are single end or paired end, but because baseParser uses _1_metric or _2_metric or _metric to determine whether samples are paired or single end, and bioflows picardTools output is always of the form _metric everything gets parsed as single end data. Also, self.library_read_type will be going away per #10. The fix should just parse all the rows without trying to decide if data is paired or single end, and add a column for read_id that has blank values.
aguang
commented
2 years ago
If alignments for
picardToolsare from paired end data, then it will currently get parsed asThis is the wrong behavior, because those files have the rows
FIRST_OF_PAIRandSECOND_OF_PAIRas well, which should have also been added to the JSON. This is because the picardTools parser usesself.library_read_typeto determine whether reads are single end or paired end, but becausebaseParseruses_1_metricor_2_metricor_metricto determine whether samples are paired or single end, and bioflows picardTools output is always of the form_metriceverything gets parsed as single end data. Also,self.library_read_typewill be going away per #10. The fix should just parse all the rows without trying to decide if data is paired or single end, and add a column forread_idthat has blank values.