Context Navigation

consense_algo.hlp

Visit:

Last change on this file was 17742, checked in by westram, 7 years ago
fix unnecessary typos.
File size: 11.7 KB

Line
1	#Please insert up references in the next lines (line starts with keyword UP)
2	UP arb.hlp
3	UP glossary.hlp
4
5	#Please insert subtopic references (line starts with keyword SUB)
6	SUB consense_tree.hlp
7	SUB bootstrap.hlp
8
9	# Hypertext links in helptext can be added like this: LINK{ref.hlp\|http://add\|bla@domain}
10
11	#*********** Title of helpfile !! and start of real helpfile ******
12	TITLE Algorithm used for consensus tree
13
14	OCCURRENCE ARB_DIST/NJ bootstrap
15	ARB_NT/Tree/Build consensus tree
16
17	DESCRIPTION ARB has its own library for calculating consensus trees, which is used
18	when
19	- calculating bootstrap trees with NJ (ARB_DIST),
20	- directly calculating consensus trees using
21	ARB_NT/Tree/Build consensus tree and
22	- from the command line tool arb_consensus_tree.
23
24	The algorithm works as follows:
25
26	- all trees are read and all occurring branches are stored in a branch-pool
27	- the consensus tree is constructed iteratively by
28	- picking the "best" branch and adding it to the tree.
29	- deleting all now impossible branches from the branch pool.
30
31	The best branch (at each time of the iteration) is determined by the following
32	criteria (listed in order of significance):
33
34	- inner branches are picked before leaf branches
35	- branches occurring more often (i.e. branches with higher bootstrap values) are picked first
36	- branches nearer to the center of the tree are preferred over more distant branches
37	- longer branches are preferred over shorter branches
38
39	If all of the above criteria are really equal for two branches, the pick-order
40	depends on their appearance in the source trees (to make results reproducible).
41
42	The distance of each branch to the center of the tree is defined by the difference
43	between the number of species on each side of the branch. Branches with an equal
44	number of species on each side are considered "at the center" of the tree.
45
46	SECTION Partial trees
47
48	If a tree doesn't contain the full set of species (i.e. if the tree is partial),
49	all branches of that tree differ from branches of full trees.
50
51	To allow merging such trees with other trees, the set of missing species is added
52	once to each side of each branch, adding two hypothetical branches to the branch pool
53	(hypothetical in that way, that they do not necessarily occur in any tree).
54
55	These two hypothetical branches are added with different probabilities:
56
57	* adding missing species to the bigger side of a branch is done with a probability of 1.0
58	* adding missing species to the smaller side of a branch is done with a
59	probability P calculated as follows:
60
61	P = ( S / B ) ^ M
62
63	where
64
65	S = number of edges on the SMALL side of the branch + 1
66	B = number of edges on the BIG side of the branch + 1
67	M = number of missing species
68
69	That probability tends towards zero for leaf branches and towards 1.0 for
70	center branches.
71
72	If you merge nearly identical topologies, which only differ by some species
73	removed from some of the topologies, the resulting bootstrap values reflect
74	the number of trees the species occurred in (e.g. 67% if species occurred
75	in 2 of 3 trees).
76
77	Another kind of branches is completely artificial: a branch with the missing
78	species on one side and all species from the partial tree on the other side.
79	These artificial branches will ONLY be considered for the output topology, if no
80	other tree defines such a branch and if it's absolutely necessary to build a
81	connected topology.
82
83	Since nothing is known about the length of such an artificial branch, a distance
84	of 1.0 will be assumed (e.g. happens when you merge disjunct trees) and a zero
85	bootstrap value will be written to the output tree.
86
87
88	SECTION Branch lengths
89
90	The branch lengths of the generated consensus tree are averaged from the input
91	branch lengths. When merging full trees, they should be quite accurate.
92
93	When merging partial trees, the lengths will be hypothetical, because all branches
94	from partial trees are hypothetical and nothing is known about how the length of
95	each branch will change by really adding the set of missing species when you use
96	a tree reconstruction tool.
97
98	As described above, hypothetical branches are inserted with a lowered
99	weight (probability) in case the missing species are estimated on the less probable
100	side of that branch. In that case the branch length of the hypothetical branch is
101	only added weighted proportionally to the assumed probability.
102
103	Example:
104
105	Imagine merging 3 trees, where species C is missing in one tree and length of
106	the branch 'AB---CD' is 0.3 in the two full trees and the length of the
107	branch 'AB---D' in the partial tree is 0.6.
108
109	The probability of adding C to the D-side of 'AB---D' is P=2/3.
110	The summarized weight for that branch is 8/3, the summarized length will
111	be 20.3 + 2/30.6 = 1.0. Then the average length calculates
112	to 1.0 / (8/3) = 0.375, i.e a bit more than 0.3, but far from 0.6.
113
114
115	NOTES Please consider using a tree reconstruction tool (e.g ARB_PARSIMONY) to recalculate
116	the branch lengths of the generated consensus tree.
117
118	EXAMPLES Example for 2 trees:
119
120
121	A C A D
122	\ / \ /
123	+--+ +--+
124	/ \ / \
125	B +--E B +--F
126	\| \|
127	D E
128
129
130	These trees contain the following branches:
131
132	A---BCDE A---BDEF
133	B---ACDE B---ADEF
134	C---ABDE D---ABEF
135	D---ABCE E---ABDF
136	E---ABCD F---ABDE
137	AB---CDE AB---DEF
138	ABC---DE ABD---EF
139
140	Consensus tree is build upon the following branch pool:
141
142	A --- BCDE[F] A[F] --- BCDE
143	B --- ACDE[F] B[F] --- ACDE
144	C --- ABDE[F] C[F] --- ABDE
145	D --- ABCE[F] D[F] --- ABCE
146	E --- ABCD[F] E[F] --- ABCD
147	A --- B[C]DEF A[C] --- BDEF
148	B --- A[C]DEF B[C] --- ADEF
149	D --- A[C]BEF D[C] --- ABEF
150	E --- A[C]BDF E[C] --- ABDF
151	F --- A[C]BDE F[C] --- ABDE
152	AB --- CDE[F] AB[F] --- CDE
153	ABC[F] --- DE ABC --- DE[F]
154	AB --- [C]DEF AB[C] --- DEF
155	AB[C]D --- EF ABD --- [C]EF
156
157	(added missing species are shown in brackets; in left column species were added to
158	more likely side; in right column to less likely side)
159
160	Bootstrap values are calculated for branches:
161
162	AB --- CDEF 100% #1
163	EF --- ABCD 56.2% #2
164	DE --- ABCF 50% #3
165	ABC --- DEF 33.3% #4
166	ABF --- CDE 16.7% #5
167	ABD --- CEF 16.7% #6
168	ABDE --- CF 12.5% #7
169	AF --- BCDE 6.25% #8
170	ACDE --- BF 6.25% #9
171	ABCE --- DF 6.25% #10
172	AC --- BDEF 6.25% #11
173	ADEF --- BC 6.25% #12
174	ABDF --- CE 6.25% #13
175	ABEF --- CD 6.25% #14
176	A --- BCDEF 100% #15
177	B --- ACDEF 100% #16
178	D --- ABCEF 100% #17
179	E --- ABCDF 100% #18
180	C --- ABDEF 50% #19
181	F --- ACBDE 50% #20
182
183	Branches were listed in the order they will be picked: first picking the
184	inner branches (#1 .. #14) in order of their bootstrap values, then the
185	leaf branches (other criteria not shown in this example).
186
187	That results in the following tree building steps:
188
189	AB---CDEF (add branch #1)
190
191
192	CD
193	/
194	/
195	AB--+ (add branch #2)
196	\56%
197	\
198	EF
199
200
201	Now it's impossible to insert branch #3 => branch is dropped!
202
203
204	C
205	/
206	/
207	AB--+ (add branch #4)
208	\33%
209	\
210	+---D
211	/
212	/56%
213	EF
214
215
216	Branches #5 to #14 are now impossible to insert => drop them
217
218	Now only leaf branches (#15 to #20) have to be added, which leads to
219	the following final topology:
220
221	A C
222	\ /
223	\ /
224	+---+
225	/ \33%
226	/ \
227	B +---D
228	/
229	/56%
230	E---+
231	\|
232	\|
233	F
234
235
236	WARNINGS None
237
238	BUGS No bugs known

Note: See TracBrowser for help on using the repository browser.

Download in other formats:

Original Format