Topology testing of phylogenies using least squares methods

Table 2 Confidence sets of trees derived from eight-taxon data sets obtained from the EMBL-ALIGN database.

Data set number	EMBL-ALIGN accession number	Sequences included	length	ΔAIC for F84	% of unresolved quartets	Number of trees in the 0.95 confidence set
						SH	ELW	GLS	WLS
1	ALIGN_000002	1,2,3,4,5,6,7,8	1632	6.4141	22.9	141	14	9	135
2	ALIGN_000205	2,3,4,6,8,10,11,12	1386	6.2104	4.3	15	6	18	9
3	ALIGN_000297	2,3,4,6,15,16,17,19	1167	0.0000	31.4	315	258	315	315
4	ALIGN_000397	2,3,4,6,7,8,9,10	1662	0.0000	24.3	2745	328	10395	10206
5	ALIGN_000398	1,2,3,4,5,6,7,8	1656	3.1567	0.0	477	20	815	77
6	ALIGN_000521	1,2,3,4,5,6,7,8	1325	8.4751	5.7	135	11	105	107
7	ALIGN_000623	2,3,4,5,6,10,11,12	1312	0.0000	0.0	380	20	10395	33
8	ALIGN_000628	2,3,4,5,7,13,17,31	1385	0.3071	0.0	141	5	117	21
9	ALIGN_000767	2,3,4,5,6,7,8,10	1386	6.6421	4.3	15	6	9	9
10	ALIGN_000771	1,2,3,4,5,6,7,8	4547	1.8389	14.3	81	9	10	15
11	ALIGN_000788	2,3,4,5,6,7,12,14	1629	0.7085	24.3	945	80	225	135
12	ALIGN_000832	2,3,4,5,6,7,9,10	1185	0.0000	10.0	327	50	49	315
13	ALIGN_000853	1,2,3,4,5,6,7,8	5307	3.3574	12.9	225	20	225	45
14	ALIGN_000930	2,3,4,5,6,7,8,9	1321	0.0000	78.6	10395	8925	10395	10393
15	ALIGN_000931	2,3,5,14,15,16,19,21	1231	0.8196	100.0	10395	9876	10395	10395
16	ALIGN_000984	2,3,4,6,7,11,12,13	1139	2.0933	45.7	10395	2344	10395	10391

The sequences taken from each alignment are listed in the second column. ΔAIC values for the Felsenstein84 nucleotide substitution model were calculated using PAUP* and ModelTest. The percentage of four-taxon subset for which the star topology was the ML solution (unresolved quartets) was calculated using TreePuzzle. The last four columns show number of trees out of possible 10395 included in the 0.95 confidence set using: expected likelihood weights (ELW) test, Shimodaira-Hasegawa (SH) test, generalised least squares (GLS) test, and weighted least squares test (WLS).

ISSN: 2730-7182