@@ -8,6 +8,29 @@ Briefly, `download your reference database <https://www.bacpop.org/poppunk/>`__
88 poppunk_assign --db database --query qfile.txt \
99 --output poppunk_clusters --threads 8
1010
11+ Nomenclature
12+ ------------
13+
14+ PopPUNK clusters are numbered from one upwards, in decreasing order of size in the initial
15+ dataset.
16+
17+ ``poppunk_assign `` will assign your genomes into these existing clusters, with the same labels as the
18+ initial run. So cluster labels, when used as documented, **do not change **.
19+
20+ In some cases, due to undersampling of the initial dataset or emergence
21+ of hybrids, some clusters may be merged. These merged clusters will be named with
22+ underscores separating the older clusters they were merges of. Use ``--external-clustering ``
23+ if you prefer other nicknames for these.
24+
25+ If you require 'stable nomenclature' where clusters never merge, use the ``--stable `` option
26+ with ``poppunk_assign ``. Each query will be assigned based on its nearest neighbour's cluster,
27+ though novel clusters will still be separately identified as 'NA'.
28+
29+ Note that maintaining stable nomenclature in a dynamic population is not possible (for any
30+ nomenclature). If you are maintaining a database and want to add new queries in, you will
31+ need to use ``--update-db `` which may merge clusters. There is no way with two or more updates
32+ of giving consistent new names to merged clusters.
33+
1134Downloading a database
1235----------------------
1336Current PopPUNK databases can be found here: https://www.bacpop.org/poppunk/
0 commit comments