Infer presence and mode of polyploidy

Below are the inputs, commands, and outputs to do several analyses with GRAMPA. The inputs are based on simulated data. For more detailed info on the simulations check our paper.

Allopolyploidy

Inputs

1000 gene trees were simulated with gain and loss using JPrIME based on the following allopolyploid-like MUL-tree:

In this scenario, lineages B and C hybridized to form an allopolyploid lineage that diversified into the x,y,z clade.

We then remove one polyploid clade from the MUL-tree to get a singly-labeled tree as input for GRAMPA. This is the same topology as above, except that the x,y,z clade sister to C is removed. This is the type of tree that typical phylogenetic reconstruction programs would produce even in the presence of allopolyploidy:

So, the input files for this search are:

1. Singly-labeled species tree: spec_tree_3a.tre
2. 1000 gene trees simulated from an allopolyploid MUL-tree: gene_trees_3a.txt

GRAMPA command

Since in reality we wouldn't know whether there is an allo-, auto-, or no polyploidy in this tree, we want GRAMPA to search all nodes as possible polyploid lineages. That means we don't specify -h1 or -h2.

grampa -s spec_tree_3a.tre -g gene_trees_3a.txt -o allo-example-output -f allo-test

Outputs

The above command would create the directory allo-example-output with six output files

• allo-test-checknums.txt
• allo-test-detailed.txt
• allo-test-dup-counts.txt
• allo-test.log
• allo-test-scores.txt
• allo-test-trees-filtered.txt

Since we are trying to determine the mode of polyploidy, we are interested in the allo-test-scores.txt file. This file contains the total reconciliation scores for each MUL-tree considered, sorted from lowest scoring tree to highest scoring tree, and looks something like this:

mul.tree        h1.node h2.node score   labeled.tree
105     <2>     C       5018    ((((((x+,y+)<1>,z+)<2>,B)<3>,A)<4>,(C,((x*,y*)<5>,z*)<6>)<7>)<8>,D)<9>
113     <3>     C       5900    ((((((x+,y+)<1>,z+)<2>,B+)<3>,A)<4>,(C,(((x*,y*)<5>,z*)<6>,B*)<7>)<8>)<9>,D)<10>
119     <4>     C       6824    ((((((x+,y+)<1>,z+)<2>,B+)<3>,A+)<4>,(C,((((x*,y*)<5>,z*)<6>,B*)<7>,A*)<8>)<9>)<10>,D)<11>
122     <4>     <5>     6824    (((((((x+,y+)<1>,z+)<2>,B+)<3>,A+)<4>,C)<5>,((((x*,y*)<6>,z*)<7>,B*)<8>,A*)<9>)<10>,D)<11>
110     <2>     <5>     7306    (((((((x+,y+)<1>,z+)<2>,B)<3>,A)<4>,C)<5>,((x*,y*)<6>,z*)<7>)<8>,D)<9>
109     <2>     <4>     7435    (((((((x+,y+)<1>,z+)<2>,B)<3>,A)<4>,((x*,y*)<5>,z*)<6>)<7>,C)<8>,D)<9>
117     <3>     <5>     7616    (((((((x+,y+)<1>,z+)<2>,B+)<3>,A)<4>,C)<5>,(((x*,y*)<6>,z*)<7>,B*)<8>)<9>,D)<10>

GRAMPA tells us MUL-tree 105 is the lowest scoring tree:

Notice that this is the same topology that was used to simulate the gene-trees. GRAMPA has successfully identified an allopolyploid MUL-tree and placed the second polyploid lineage on the correct branch!

Autopolyploidy

1000 gene trees were simulated with gain and loss using JPrIME based on the following autopolyploid-like MUL-tree:

In this scenario, a lineage sister to species C underwent autopolyploidization and subsequently diversified into the x,y,z clade.

We then remove one polyploid clade from the MUL-tree to get a singly-labeled tree as input for GRAMPA. This is the same topology as above, except that one x,y,z clade is removed. This is the type of tree that typical phylogenetic reconstruction programs would produce even in the presence of autopolyploidy:

So, the input files for this search are:

1. Singly-labeled species tree: spec_tree_18.tre
2. 1000 gene trees simulated from an allopolyploid MUL-tree: gene_trees_18.txt

GRAMPA command

Since in reality we wouldn't know whether there is an allo-, auto-, or no polyploidy in this tree, we want GRAMPA to search all nodes as possible polyploid lineages. That means we don't specify -h1 or -h2.

grampa -s spec_tree_18.tre -g gene_trees_18.txt -o auto_example_output -f auto_test -v 0

Outputs

The above command would create the directory auto_example_output with six output files

• auto-test-checknums.txt
• auto-test-detailed.txt
• auto-test-dup-counts.txt
• auto-test.log
• auto-test-scores.txt
• auto-test-trees-filtered.txt

Since we are trying to determine the mode of polyploidy, we are interested in the auto-test-scores.txt file. This file contains the total reconciliation scores for each MUL-tree considered, sorted from lowest scoring tree to highest scoring tree, and looks something like this:

mul.tree        h1.node h2.node score   labeled.tree
119     <3>     <3>     4807    (((B,A)<1>,((((x+,y+)<2>,z+)<3>,((x*,y*)<4>,z*)<5>)<6>,C)<7>)<8>,D)<9>
0       NA      NA      5476    (((B,A)<1>,(((x,y)<2>,z)<3>,C)<4>)<5>,D)<6>
85      D       D       5493    (((B,A)<1>,(((x,y)<2>,z)<3>,C)<4>)<5>,(D+,D*)<6>)<7>
105     <2>     z       5776    (((B,A)<1>,(((x+,y+)<2>,(z,(x*,y*)<3>)<4>)<5>,C)<6>)<7>,D)<8>
110     <2>     <3>     5776    (((B,A)<1>,((((x+,y+)<2>,z)<3>,(x*,y*)<4>)<5>,C)<6>)<7>,D)<8>
71      C       C       5812    (((B,A)<1>,(((x,y)<2>,z)<3>,(C+,C*)<4>)<5>)<6>,D)<7>
97      <1>     <1>     5817    ((((B+,A+)<1>,(B*,A*)<2>)<3>,(((x,y)<4>,z)<5>,C)<6>)<7>,D)<8>

GRAMPA tells us MUL-tree 119 is the lowest scoring tree:

Notice that this is the same topology that was used to simulate the gene-trees, and that the h1 and h2 nodes of this MUL tree are the same, indicating autopolyploidy. GRAMPA has successfully identified an autopolyploid MUL-tree on the correct branch!

No polyploidy

1000 gene trees were simulated with gain and loss JPrIME based on the following singly-labeled tree:

In this scenario, no polyploidy has occurred and this is the same tree we give to GRAMPA.

So, the input files for this search are:

1. Singly-labeled species tree: spec_tree_33.tre
2. 1000 gene trees simulated from an allopolyploid MUL-tree: gene_trees_33.txt

GRAMPA command

Since in reality we wouldn't know whether there is an allo-, auto-, or no polyploidy in this tree, we want GRAMPA to search all nodes as possible polyploid lineages. That means we don't specify -h1 or -h2.

grampa -s spec_tree_33.tre -g gene_trees_33.txt -o nop_example_output -f nop_test -v 0

Outputs

The above command would create the directory nop_example_output with four output files

• nop-test-checknums.txt
• nop-test-detailed.txt
• nop-test-dup-counts.txt
• nop-test.log
• nop-test-scores.txt
• nop-test-trees-filtered.txt

Since we are trying to determine the mode of polyploidy, we are interested in the nop-test-scores.txt file. This file contains the total reconciliation scores for each MUL-tree considered, sorted from lowest scoring tree to highest scoring tree, and looks something like this:

mul.tree        h1.node h2.node score   labeled.tree
0       NA      NA      4115    ((((((x,y)<1>,z)<2>,B)<3>,A)<4>,C)<5>,D)<6>
85      D       D       4217    ((((((x,y)<1>,z)<2>,B)<3>,A)<4>,C)<5>,(D+,D*)<6>)<7>
71      C       C       4370    ((((((x,y)<1>,z)<2>,B)<3>,A)<4>,(C+,C*)<5>)<6>,D)<7>
57      A       A       4423    ((((((x,y)<1>,z)<2>,B)<3>,(A+,A*)<4>)<5>,C)<6>,D)<7>
43      B       B       4554    ((((((x,y)<1>,z)<2>,(B+,B*)<3>)<4>,A)<5>,C)<6>,D)<7>
120     <4>     D       4558    ((((((x+,y+)<1>,z+)<2>,B+)<3>,A+)<4>,C)<5>,(D,((((x*,y*)<6>,z*)<7>,B*)<8>,A*)<9>)<10>)<11>
113     <3>     C       4632    ((((((x+,y+)<1>,z+)<2>,B+)<3>,A)<4>,(C,(((x*,y*)<5>,z*)<6>,B*)<7>)<8>)<9>,D)<10>

GRAMPA tells us that the input species tree (i.e. the singly-labeled tree) is the lowest scoring tree:

Notice that this is the same topology that was used to simulate the gene-trees and contains no polyploidy events. GRAMPA has successfully determined that no polyploidy has occurred among these lineages!