Skip to content

Latest commit

 

History

History
124 lines (109 loc) · 5.92 KB

README.md

File metadata and controls

124 lines (109 loc) · 5.92 KB
Raw

TREE-QMC-BP Tutorial

  1. Clone the TREE-QMC Github repository and build TREE-QMC following the instructions in the README.

  2. Go to tutorial directory.

cd tutorial/characters
  1. Run TREE-QMC-BP on the example input data. The nexus file contains presence/absence patterns for CR1 retrotransposon insertions from Cloutier et al., Syst Biol, 2019, plus the 44 additional characters found by Simmons et al., Mol Phy Evol, 2022. The command for running TREE-QMC in BP mode is
../../tree-qmc \
    --bp \
    --root galGal \
    --support \
    -i 4345ratites.nex \
    -o treeqmc-bp-4345ratites.tre

This command builds a species tree from the quartets induced by each character; see Springer et al., J Heredity, 2020. If both --support and --bp options are used, branch lengths will be computed using the MLE method described in Molloy, Gatesy, Springer, Syst Biol, 2022.

IMPORTANT: BP mode is for low-homoplasy bilallelic character matrices; however, you can run these analyses for multi-allelic characters (e.g., nucleotides) by replacing --bp with --chars in the command above.

TREE-QMC-BP recovers the same species tree topology as Cloutier et al., Syst Biol, 2019, who estimated the species tree from gene trees using MP-EST (see the model tree below).

model

The placement of Rhea in this species tree is debated. The quartet suport for this branch is 'q1=0.499331 (with f1=18.650000 and EN=37.350000), which means that about half of the quartets around the focal branch support Rhea being sister to Kiwi+Emu+Cassowary. However, the low EN suggests that there is limited signal in the character matrix for resolving this branch. The --pcsonly flag can be used to explore this futher by providing the quartet support for the placement of Rhea for each character.

  1. The goal of Partitioned Coalescence Support (PCS), described by Gatesy et al., Mol Phy Evol, 2019, is to evaluate the quartet support of each character (or gene tree) for resolving a focal branch in the species tree. TREE-QMC can be used to compute PCS by providing an additional input: a species tree with PCS flagging the focal branch; see species-tree-for-pcs.tre and unresolved-species-tree-for-pcs.tre. The command for running TREE-QMC in PCS mode is
../../tree-qmc \
    --bp \
    --pcsonly species-tree-for-pcs.tre \
    -i 4345ratites.nex \
    -o pcs-bp-4345ratites.tsv

The output shown of TREE-QMC is below (plus our annotations indicated by #). The four possible trees related to placement of Rhea are

  • t1 = A,B|C,D = Rhea,Kiwi+Emu+Cassowary|Tinamou,Chicken+Ostrich (matches input species trees)
  • t2 = A,C|B,D = Rhea,Tinamou|Kiwi+Emu+Cassowary,Chicken+Ostrich
  • t3 = Rhea,Chicken+Ostrich|Kiwi+Emu+Cassowary,Tinamou
  • t4 = A,C,C,D = Rhea,Kiwi+Emu+Cassowary,Tinamou,Chicken+Ostrich (no quartet / polytomy)

Thus, there are at most |x| * |y| * |z| * |w| = 2 * 5 * 4 * 2 = 80 quartets that inform the resolution of the focal branch (although a character can induce fewer quartets if there are missing data). Because each character can induce multiple quartets, it can split its vote between t1, t2, t3, or t4. Overall, the PCS analysis finds 61 characters (out of 4345 characters) support at least one of t1, t2, and t3. If polytomies (t4) are ignored, 31 characters (51%) support only t1, 14 characters (23%) support t2, and 16 characters (26%) support t3 (note the number of quartets contributed by each of these characters can differ). This PCS analysis is close to the estimates above for the branch (note f1 above is computed by dividing column f_xy|zw by column totalf and then summing the result).

# x = rheAme,rhePen
# y = aptHaa,aptOwe,aptRow,casCas,droNov
# z = cryCin,tinGut,eudEle,notPer
# w = strCam,galGal
id	position	f_xy|zw	f_xz|yw	f_xw|yz	f_xyzw	totalf
3340	3341	30	0	0	0	30 # t1
3341	3342	40	0	0	0	40 # t1
3342	3343	40	0	0	0	40 # t1
3343	3344	40	0	0	0	40 # t1
3344	3345	30	0	0	0	30 # t1
4276	4277	24	0	0	16	40 # t1
4277	4278	24	0	0	16	40 # t1
4278	4279	24	0	0	16	40 # t1
4279	4280	18	0	0	12	30 # t1
4280	4281	12	0	0	8	20 # t1
4281	4282	16	0	0	24	40 # t1
4282	4283	8	0	0	12	20 # t1
4283	4284	16	0	0	24	40 # t1
4284	4285	12	0	0	18	30 # t1
4285	4286	16	0	0	24	40 # t1
4286	4287	16	0	0	24	40 # t1
4296	4297	0	4	0	6	10 # t2
4297	4298	0	0	24	16	40 # t3
4298	4299	0	0	16	24	40 # t3
4299	4300	0	40	0	0	40 # t2
4300	4301	0	10	0	10	20 # t2
4301	4302	0	0	30	0	30 # t3
4302	4303	0	0	30	0	30 # t3
4303	4304	0	0	40	0	40 # t3
4304	4305	0	0	30	0	30 # t3
4305	4306	0	0	40	0	40 # t3
4306	4307	0	0	40	0	40 # t3
4307	4308	40	0	0	0	40 # t1
4308	4309	40	0	0	0	40 # t1
4309	4310	0	40	0	0	40 # t2
4310	4311	0	40	0	0	40 # t2
4311	4312	0	0	16	24	40 # t3
4312	4313	0	0	8	12	20 # t3
4313	4314	0	0	12	18	30 # t3
4314	4315	0	0	16	24	40 # t3
4315	4316	0	0	12	18	30 # t3
4316	4317	0	0	24	16	40 # t3
4317	4318	0	0	18	12	30 # t3
4318	4319	0	0	18	12	30 # t3
4319	4320	24	0	0	16	40 # t1
4320	4321	24	0	0	16	40 # t1
4321	4322	24	0	0	16	40 # t1
4322	4323	24	0	0	16	40 # t1
4323	4324	24	0	0	16	40 # t1
4324	4325	18	0	0	12	30 # t1
4325	4326	16	0	0	24	40 # t1
4326	4327	16	0	0	24	40 # t1
4327	4328	12	0	0	18	30 # t1
4328	4329	16	0	0	24	40 # t1
4329	4330	8	0	0	24	32 # t1
4330	4331	16	0	0	24	40 # t1
4331	4332	16	0	0	24	40 # t1
4332	4333	0	24	0	16	40 # t2
4333	4334	0	18	0	12	30 # t2
4334	4335	0	16	0	24	40 # t2
4335	4336	0	12	0	18	30 # t2
4336	4337	0	16	0	24	40 # t2
4337	4338	0	12	0	18	30 # t2
4338	4339	0	16	0	24	40 # t2
4339	4340	0	12	0	18	30 # t2
4340	4341	0	16	0	24	40 # t2