Differences

This shows you the differences between two versions of the page.

--- phylogeny_with_beast [2013/06/29 15:28]
anniearchambault
+++ phylogeny_with_beast [2013/09/11 17:30] (current)
anniearchambault
@@ Line 1: / Line 1: @@
-====Testing evolutionary hypothesis with BEAST2.0====
+====Dating a node with BEAST2.0====
 For details, read http://beast2.cs.auckland.ac.nz/index.php/Main_Page and http://beast2.cs.auckland.ac.nz/index.php/FAQ .
+For any problem, do not hesitate to browse through the list of questions on the BEAST forum https://groups.google.com/forum/#!forum/beast-users .
 ===Exercise description===
@@ Line 13: / Line 14: @@
 http://qcbs.ca/wp-content/uploads/2013/06/ProteaFaurea_trnL.txt
-The files include sequences of the ITS and the trnL regions, from 74 species of the //Protea// genus, which comprises a total of 115 species, and the same loci from one specimen of //Faurea//, the sister genus of //Protea//. According to another study, the split leading to  //Faurea// and to //Protea// is 28.4 Ma (central 95% range 24.4–32.3 Ma). With an approximate date for this single root node, and the DNA sequences in hand, we will try to answer the question: How old is one of the oldest clade in //Protea//, comprising //P. cryophila, P. lorea, P. pruinosa, P. scabriuscula, P. scolopendrifolia// ?
+The files include sequences of the ITS and the trnL regions, from 74 species of the //Protea// genus, which comprises a total of 115 species, and the same loci from one specimen of //Faurea//, the sister genus of //Protea//. According to another study, the split leading to  //Faurea// and to //Protea// is 28.4 Ma (central 95% range 24.4–32.3 Ma). With an approximate date for this single root node, and the DNA sequences in hand, we will try to answer the question: How old is one of the oldest clade in //Protea//, comprising //P. cryophila, P. pruinosa, P. scabriuscula, P. scolopendrifolia// ?
@@ Line 21: / Line 22: @@
 ===Step by step instructions===
-(Optional, if time allows) Before importing into BEATi, use any phylogenetic program you liked to run an analysis of the sequence datasets provided, to confirm that the 5 species listed above truly group into a clade.
+(Optional, if time allows) Before importing into BEATi, use any phylogenetic program you liked, to run an analysis of the sequence datasets provided, to confirm that the 5 species listed above truly group into a clade.
-**1** Import the two files in BEAUti, with File/Import Alignment. For each, we will now go over the top panel (Partitions / Tip dates / Site Model / Clock Model / Priors / MCMC), to choose all the options required to answer the question. BEAUti will then create a .xml file readable by the BEAST program.
+**1** Import the two files in BEAUti, with File/Import Alignment. For each, we will now go over the top panel (Partitions / Tip dates / Site Model / Clock Model / Priors / MCMC), to choose all the required options. BEAUti will then create a .xml file readable by the BEAST program.
 **2** Tip dates: Nothing is added or modified to this panel
@@ Line 37: / Line 38: @@
 {{:beast_subs_model1.png?400|Substitution models in BEAUti for ITS}}
-For the trnL, we will choose the model selected by JModelTes2 \\
+For the trnL, we will choose the model selected by JModelTest2 \\
+{{:jmodeltest_seeresults.png?200|best substitution models from jModelTest for trnL}}
 {{:beast_subs_model_trnl.png?400|Substitution models in BEAUti for trnL}}
@@ Line 45: / Line 47: @@
 **5** Setting the Priors \\
-This step is the most complicated of the whole process.
+This step is the **most critical and most complicated** of the whole process.
 For the Tree prior, we will select the Yule model. Yule models are generally appropriate with sequences from different species, while Coalescent models are for different populations of a same species. Do the same for both partitions (ITS and trnL)
@@ Line 55: / Line 57: @@
 For the GammaShape, leave it as it is, Exponential, Mean=1, check estimate.
-For kappa for ITS, also leave it as it is, LogNormal, M=1, S=1.25
+For kappa for ITS, also leave it as is, LogNormal, M=1, S=1.25
 We will now add two calibration nodes, to answer the original question. One node will be the root, it will include all taxa (including the single Faurea), and we will set a specific known age to it. The other node will include the 5 species listed above, and we will name this clade "OldClade_ITS" and "OldCladetrnL". Beast will estimate an age for this clade. \\
@@ Line 63: / Line 65: @@
 For the two created sets, check the monophyletic box, select Normal, Mean=28.4, Sigma=2.5, do not check the estimate box.
-Create two new taxon sets, for the trnL and the ITS tree, each containing the five species //P. cryophila, P. lorea, P. pruinosa, P. scabriuscula, P. scolopendrifolia//. Label the sets OldClade_ITS and OldClade_trnL. For each set, Select Normal, check the monophyletic box, and add a reasonable number for Mean, Mean=11, check the estimate box, Sigma=1, check estimate. \\
+Create two new taxon sets, for the trnL and the ITS tree, each containing the four species //P. cryophila, P. pruinosa, P. scabriuscula, P. scolopendrifolia//. Label the sets OldClade_ITS and OldClade_trnL. For each set, Select Normal, check the monophyletic box, and add a reasonable number for Mean, Mean=11, check the estimate box, Sigma=1, check estimate. \\
 {{:beast_taxaset_normal_1.png?400|}}
-You can notice that new parameters appeared. Leave them as is for now.
+You can notice that new parameters appeared. Leave them as for the exercise.
 **6** MCMC \\
@@ Line 83: / Line 85: @@
 **8** Run the analysis \\
 Save your the BEAUti settings as a .xml file. Open the BEAST program, Load the .xml file, and run the search. When completed, modify the name. This is because for you be able to re-launch an BEAST analysis with the same .xml file.
+++++ Or use this example file  |
+Here is one xml file: http://qcbs.ca/wp-content/uploads/2013/09/ProteaFaurea_ITS_trnl_test18.xml_.txt
+++++
 **9** Analyze the results \\
-Open the program Tracer. Load the newly created .log file. Or load the  \\
+Open the program Tracer. Load the newly created .log file.  \\
 Select the two mrcatime(OldClade) values, and look at the Estimates. \\
+Try answer the questions http://qcbs.ca/wp-content/uploads/2013/09/Beast_questions3.pdf  \\
 {{:tracer_mrca_old.png?400|}}
 **10** Look at the trees \\
 Open the TreeAnnotator program. \\
-Burnin (in nb of trees)= 500 \\
+Burnin (in nb of trees)= 30000 \\
-Posterior probability limit=0.5 (a value of 0 will annotate all trees) \\
+Posterior probability limit=0.9 (a value of 0 will annotate all trees) \\
 Target Tree Type= Maximum clade credibility tree \\
 Node height= Mean heights \\
 Choose the trnL input file and give a name for output file. \\
+{{:treeannotator_screen.png?400|}}
 Open the FigTree program  \\
@@ Line 101: / Line 108: @@
 Re-root at Faurea, and order nodes, decreasing or increasing at your preference \\
 Check "Node Bars", select "Height_95%_HPD" \\
+{{:figtree_trnl2.png?400|}}