1. 20 Jul, 2018 2 commits
  2. 12 Jul, 2018 2 commits
    • Mikael Salson's avatar
      data/segmentation-2.fa: Small change in random sequence · 18ce10f2
      Mikael Salson authored
      That random sequence was supposed to have too few v/j information (because
      it's random!). However, because of bad luck there was some tiny similarities
      with IGH but the sequence was not segmented due to inconsistencies with the
      strand.
      This is still cool because the random sequence is not segmented. However here
      we just want to test that a sequence with too few v/j is reported as such. As
      the strand consistency is tested first in our algorithm, we didn't report the
      too few v/j information. Therefore I modified a single nucleotide in the
      sequence to get rid of this strand inconsistency.
      
      I think I always feel guilty when modifying a test example. That's probably
      why I needed to make such a long comment.
      
      See !148
      18ce10f2
    • Mikael Salson's avatar
      data/chimera-fake-half.fa: cgcgcgcg is not a good example · f5c1d159
      Mikael Salson authored
      It appears in TRBJ2-5*01 which leads to false positives.
      The purpose of our dimer was only to have some rubbish sequence.
      Thus I have no qualms about changing it.
      
      Désolé pour le côté scatologique du dimer.
      
      See MR !148
      f5c1d159
  3. 11 Jul, 2018 1 commit
  4. 07 Jul, 2018 1 commit
    • Mikael Salson's avatar
      algo/tests: Bad filtering with -Z 1 · 30c16f63
      Mikael Salson authored
      Sequece aligned with -Z 1 is IGHV4-59*06 while it should be IGHV4-31*0[23].
      
      IGHV4-59*06 is a pretty good choice for this sequence as it has 100% identity
      however it is 4nt shorter compared to IGHV4-31*0[23] (which also have 100% identity).
      
      Example for #3344.
      30c16f63
  5. 05 Jul, 2018 1 commit
  6. 13 Jun, 2018 1 commit
  7. 04 Jun, 2018 1 commit
  8. 25 May, 2018 2 commits
  9. 01 Feb, 2018 2 commits
  10. 19 Jan, 2018 1 commit
  11. 18 Jan, 2018 1 commit
  12. 17 Jan, 2018 2 commits
  13. 16 Jan, 2018 3 commits
    • Mikael Salson's avatar
      segment.h: The minimal window length was actually quite small. · 8aa67229
      Mikael Salson authored
      It was hard to make some tests (eg. unsegmentation-causes.should-get) have a
      too short window. Modifying a constant just so that the test passes is bad
      practice, but this is not the case here (at least I believe so).  Having
      difficulties to have a too short w also meant that in practice we would almost
      never see a “too short w” again. This could be nice. But we know that the
      position of the window is not accurate (it is not designed to be). Therefore
      20nt, if they are perfectly centered may be ok. But we know that it may not be
      well centered (and we even allow some shifts). Therefore 30nt seems to be a
      safer limit.
      
      See #2913
      8aa67229
    • Mikael Salson's avatar
      testSegment: Sequence was now segmented but we wanted to test the opposite · 577df6b1
      Mikael Salson authored
      The sequence has been shortened and the tests updated accordingly.
      577df6b1
    • Mikael Salson's avatar
      bug2910.should-get: Do not rely on .should-vdj file · 919ddd98
      Mikael Salson authored
      The should-vdj test fails for reasons unrelated with this issue.
      Thus we have to get rid of this should-vdj test (see #2767 for that matter)
      and only test what we want here: that the clone is properly segmented
      919ddd98
  14. 30 Nov, 2017 1 commit
  15. 15 Sep, 2017 3 commits