filter by XA tag

[colindaven]

The aligner bwa mem provides the XA tag for supplemental alignments.

They are present even after using an MQ30 filter, i.e. MQ is generally 30-40 (surprising!).

We could trial removing reads/alignments with these and test results. Generally, other species from the same genera are picked up (Veillonella parvula -> V. dispar, etc), but Schaalia o. to Actinomyces sp are also present.

Streptococcus is, as ever, problematic.

XA:Z:NZ_LR134375.1_Veillonella_dispar_strain_NCTC11831_chromosome_1,-2088145,9S67M,7;

/ngsssd1/rcug/wochenende_test/XA_tags

Edit, will likely have a small effect (2-3 %?) based on one example only.

eg current example CF_sm2 

278k total microbial non-human aligned reads
9k reads with XA tag