chimera.ses(method=ccode) weird windows, way high anova

[1.7.2 and 1.8 Win32+linux]

I’m running the ccode chimera-checking method with default settings and am experiencing some unexpected behavior.

  1. I’m always seeing 11 windows when the default should give 10. The last window has only 2 nucleotides in it and does not end at the last base (1133 instead of 1135)
  2. Some anova values are shockingly high (33874344.000.000) [Windows only. These values are zero in linux.]
    and my favorite:
  3. It declares that 100% of my sequences are chimeric!

sequences available upon request.

Also… I notice that mothur doesn’t stop you from checking your sequences against a database that hasn’t been lane-masked with the same *.filter file used on your sequences. This isn’t the issue in my case, but again, I wish mothur would act a bit more mothurly and nag just a little bit.

Robin