ac5d7c6726a1d4ccbea22592fa64de309869e9bf
angie
  Sun Feb 26 11:03:27 2023 -0800
Added many false reversions that were causing trouble in XBB, as well as a few for BQ.1, BQ.1.1, XBB.1(+27915) and XBB.1.5.

diff --git src/hg/utils/otto/sarscov2phylo/branchSpecificMask.yml src/hg/utils/otto/sarscov2phylo/branchSpecificMask.yml
index e21ebc9..b725f92 100644
--- src/hg/utils/otto/sarscov2phylo/branchSpecificMask.yml
+++ src/hg/utils/otto/sarscov2phylo/branchSpecificMask.yml
@@ -21,31 +21,31 @@
     # Deletions at S:157-158 (22029-22034), ORF8:119-120 (28248-28253); mask adjacent noisy bases too
     ranges: [ [ 22027, 22034], [28246, 28254] ]
     sites: [ 21302, 21304, 21305, # https://github.com/cov-lineages/pango-designation/issues/398
              21846,               # S:95 noisy -- caused split AY.100
              28461,               # https://github.com/cov-lineages/pango-designation/issues/435
              28271 ]              # deletion
 
 BA.1:
     representative: England/DHSC-CYBJ4Y8/2022
     ranges: [ [ 6513, 6515 ],
               [ 11283, 11291 ],   # https://github.com/cov-lineages/pango-designation/issues/361
               [ 21765, 21770 ],
               [ 21988, 21995 ],
               [ 22194, 22217 ],   # deletion 22194-22196, insertion after 22204, big mess after that
               [ 28362, 28370 ] ]
-    sites: [ 22813, 22898, 22882, 23854 ]
+    sites: [ 203, 22813, 22898, 22882, 23854 ]
 
 BA.1.1:
     # On the BA.1 branch, so it inherits BA.1 masking.
     representative: England/ALDP-2BEB0A0/2021
     ranges: [ [ 76, 77 ] ]
     reversions: [ G26530A ]       # false reversions on this really messed up the BA.1.1.1 branch.
                                   # XD should have this reversion, oh well.
 
 BA.2:
     # I'm including some 5' UTR and 3' UTR sites now; in retrospect could have started earlier.
     # I would mask 210 too but it's useful for finding breakpoints of Delta/BA.2 recombinants.
     # 212 is also noisy but I'm leaving it in as a red flag for 210 mutations that might be noise.
     representative: England/DHSC-CYBAB7G/2022
     representativeBacktrack: 1
     ranges: [ # 5' UTR
@@ -93,15 +93,54 @@
     ranges: [ [ 686, 694 ], [ 21765, 21770 ] ]
 
 BA.5:
     # BA.5 is placed on the BA.2 branch so it inherits all the BA.2 sites.
     representative: England/PHEP-YYFJPAM/2022
     ranges: [ [ 21765, 21770 ] ]
     # Some of these should be reverted in recombinants, but we're pretty much past the point of
     # simultaneous Delta/Omicron and the noise from false reversions is so intolerable that we'll
     # just have to watch out for missing reversions when working with recombinants.
     reversions: [ G670T, T2790C, T3037C, A4184G, T4321C, G9424A, T9534C, T10198C, A12160G,
                   T15714C, T17410C, G18163A, T19955C, G20055A, T21618C, G22200T, A22578G,
                   T22674C, C22679T, T22686C, G22688A, A22775G, T22813G, G22882T, G22917T,
                   A22992G, A22995C, C23013A, G23018T, G23055A, T23063A, C23075T, G23403A,
                   T23525C, G23599T, A23604C, T25000C, A26529G, G26577C, A26709G, T27807C,
                   T27889C, T28271A, C29510A ]
+
+BQ.1:
+    # Inherits from BA.5
+    # Mask some suspicious run-of-nearby-site mutations that cause a mini-BQ.1 branch and
+    # caused trouble when minimizing the tree for pango-designation v1.18.1 pangolin-data release.
+    representative: England/PHEP-YYGYEQS/2022
+    sites: [ 114, 117, 244, 256, 258, 261, 264 ]
+
+BQ.1.1:
+    # Inherits from BQ.1
+    representative: England/DHSC-CYF1KSU/2022
+    reversions: [ C22893A, A22942T ]
+
+XBB:
+    # Inherits from BA.2
+    # Lots of false reversions; I would also mask G405A except that would be wrong for XBN.
+    representative: England/PHEC-YY8GAXZ/2022
+    reversions: [ T2790C, T3037C, A4184G, T4321C, T9344C, G9424A, T10198C, C17859T, G19326A,
+                  T21618C, C21810T, A22000C, G22109C, C22577G, A22578G, C22599G, A22664C, C22679T,
+                  T22686C, G22688A, T22813G, C22895G, C22896T, A22898G, G22942T, A22992G, A22995C,
+                  C23013A, C23019T, C23031T, T24424A, A24469T, T25000C, T26060C, G26577C, A26709G,
+                  T26858C, T27807C ]
+XBB.1:
+    # Inherits from XBB
+    # Here I'm going a bit past the technical start of XBB.1, to include G27915T to be more sure
+    # that it's XBB.1 before we mask out the defining mutation of XBB.1
+    representative: England/LSPA-32578111/2022
+    reversions: [ T22317G ]
+
+XBB.1.5:
+    # Inherits from XBB.1
+    # Don't believe reversions on 27915 once we're as far as XBB.1.*
+    representative: England/BRBR-326C0A5A/2023
+    reversions: [ T27915G ]
+
+XBC:
+    # Inherits nothing!  Should find out its deletions.
+    representative: Philippines/PH-VUI-142736/2022
+    reversions: [ G5584A, T13019C, T25000C, T28271A ]