Bootstrap <1%issue

[Christian Rinke]

Hello,

I have created an IQ bootstrap tree, by bootstrapping my alignment and then running 100 trees: 

iqtree -sup      2.1.iqtree.c10.pmsf.fasttreeStart.rooted.tree    boot_trees_100x.trees

Then I concatenated all trees into one file and used it to map the BS values onto my snapshot tree:

module load iqtree

iqtree -sup      snapshot.tree    boot_trees_100x.trees

The strange thing is that there is a trail for <1% BS that goes from the root all the way to one genus (see screenshot). I am confident that most of these <1% BS values are nonsense, since I get good support using other methods.

This seems a very odd problem, but I was wondering did anyone of you ever encounter something like this?

Cheers,

Chris

[Minh Bui]

Hi Chris,

That looks strange. Before we can help, can you please paste the command line used to compute the bootstrap trees?

Can you also send the newick tree file used to generate this figure?

Did you follow this FAQ: http://www.iqtree.org/doc/Frequently-Asked-Questions#how-do-i-save-time-for-standard-bootstrap ?

Cheers,

Minh

--
You received this message because you are subscribed to the Google Groups "IQ-TREE" group.
To unsubscribe from this group and stop receiving emails from it, send an email to iqtree+un...@googlegroups.com.
To post to this group, send email to iqt...@googlegroups.com.
Visit this group at https://groups.google.com/group/iqtree.
To view this discussion on the web visit https://groups.google.com/d/msgid/iqtree/8a0b29a1-08ed-4327-ae8b-8ce6229e7ddc%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
<IQtree_artifact_BS.gif>

[Christian Rinke]

Hi Minh,

the command was (100 in total):

iqtree -s   boots/bootstrap_aln.1.faa     -nt AUTO     -m LG+C10+F+G      -ft   01.fasttree.1248x.wag.g.tree   -pre   bootstrap_aln.1_pmsf ....

iqtree -s   boots/bootstrap_aln.100.faa     -nt AUTO     -m LG+C10+F+G      -ft   01.fasttree.1248x.wag.g.tree   -pre   bootstrap_aln.100_pmsf

(sorry, I had the wrong command in my first email)

# concatenate trees into one file

cat *.treefile >  boot_trees_100x.trees      # newick file attached <boot_trees_100x.trees>   

# Create BS tree file 

module load iqtree  

iqtree -sup      2.1.iqtree.c10.pmsf.fasttreeStart.rooted.tree    boot_trees_100x.trees          # newick file attached <boot_trees_100x.trees.suptree>

# Decorate tree

# in house script,  newick file attached <3.iqtree.c10.pmsf.fasttreeStart.BS.decor.tree>     

Best,

Chris

On Monday, 8 July 2019 19:55:28 UTC+10, Minh Bui wrote:

Hi Chris,

That looks strange. Before we can help, can you please paste the command line used to compute the bootstrap trees?

Can you also send the newick tree file used to generate this figure?

Did you follow this FAQ: http://www.iqtree.org/doc/Frequently-Asked-Questions#how-do-i-save-time-for-standard-bootstrap ?

Cheers,

Minh

On 8 Jul 2019, at 11:55 am, Christian Rinke <christi...@gmail.com> wrote:

Hello,

I have created an IQ bootstrap tree, by bootstrapping my alignment and then running 100 trees: 

iqtree -sup      2.1.iqtree.c10.pmsf.fasttreeStart.rooted.tree    boot_trees_100x.trees

Then I concatenated all trees into one file and used it to map the BS values onto my snapshot tree:

module load iqtree

iqtree -sup      snapshot.tree    boot_trees_100x.trees

The strange thing is that there is a trail for <1% BS that goes from the root all the way to one genus (see screenshot). I am confident that most of these <1% BS values are nonsense, since I get good support using other methods.

This seems a very odd problem, but I was wondering did anyone of you ever encounter something like this?

Cheers,

Chris

--
You received this message because you are subscribed to the Google Groups "IQ-TREE" group.

To unsubscribe from this group and stop receiving emails from it, send an email to iqt...@googlegroups.com.

[Minh Bui]

Hi Chris,

Can you pls also send this tree: 2.1.iqtree.c10.pmsf.fasttreeStart.rooted.tree

It could be a problem in this file, or your in-house script. Everything else looks fine.

Cheers,

Minh

To unsubscribe from this group and stop receiving emails from it, send an email to iqtree+un...@googlegroups.com.

To post to this group, send email to iqt...@googlegroups.com.
Visit this group at https://groups.google.com/group/iqtree.

To view this discussion on the web visit https://groups.google.com/d/msgid/iqtree/d5f73777-9a3e-4f63-b660-4671d37514c5%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

<boot_trees_100x.trees><3.iqtree.c10.pmsf.fasttreeStart.BS.decor.tree><boot_trees_100x.trees.suptree>

[Christian Rinke]

Thanks Minh, the tree is attached.

Minh

[Minh Bui]

Alright, I could reproduce this now. 

My next question is: how did you reconstruct this tree? 

When I computed a consensus tree from 100 bootstrap trees with:

iqtree-test -con boot_trees_100x.trees

I don’t see such “trail” anymore. All branches of the consensus tree have okey supports, mostly > 50%.

So I conclude there is something wrong with the tree you used to map the support values onto.

Alternatively, you can use the consensus tree.

Cheers

Minh

To unsubscribe from this group and stop receiving emails from it, send an email to iqtree+un...@googlegroups.com.

To post to this group, send email to iqt...@googlegroups.com.
Visit this group at https://groups.google.com/group/iqtree.

To view this discussion on the web visit https://groups.google.com/d/msgid/iqtree/55509013-6df4-4516-b7e5-429024b3b748%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

<2.1.iqtree.c10.pmsf.fasttreeStart.rooted.tree>