Getting node support values in the ncov workflow

Hi - I’ve found discussion of this scattered over several help topics and GitHub issues, but am having trouble figuring out what action I should take. I am running the default ncov workflow with very minor modifications, not interested in doing extensive customization, but I would like to get bootstrap values in order to be able to annotate the nodes of my tree. (I would preferably like these to be present in the newick file itself, not in the auspice json output, since I’m not going to use auspice for my publication figures). My questions are 1) are the node support values already present somewhere in the results of running the default workflow? (Just to make sure I’m not missing something) And if not…2) how do I modify the workflow to get them?
I tried following this: Enable bootstrap support by huddlej · Pull Request #839 · nextstrain/augur · GitHub to modify my default parameters.yaml file to say --tree-builder-args '-bb 1000 -bnni -czb' and override-default-args: true but I get the following error message on the augur tree step:

        augur tree             --alignment results/filtered.fasta             --tree-builder-args '-bb 1000 -bnni -czb'             --exclude-sites defaults/sites_ignored_for_tree_topology.txt             --output results/tree_raw.nwk             --nthreads 8 2>&1 | tee logs/tree.txt

ERROR: Shell exited 2 when running: iqtree2 -nt 8 -s results/masked_filtered-delim.fasta -m GTR -ninit 2 -n 2 -me 0.05 -bb 1000 -bnni -czb > results/masked_filtered-delim.iqtree.log
Command output was:
  Ultrafast bootstrap does not work with -fast, -te or -n option

7 masking sites read from defaults/sites_ignored_for_tree_topology.txt
Building a tree via:
	iqtree2 -nt 8 -s results/masked_filtered-delim.fasta -m GTR -ninit 2 -n 2 -me 0.05 -bb 1000 -bnni -czb > results/masked_filtered-delim.iqtree.log
	Nguyen et al: IQ-TREE: A fast and effective stochastic algorithm for estimating maximum likelihood phylogenies.
	Mol. Biol. Evol., 32:268-274.

Please see the log file for more details: results/masked_filtered-delim.iqtree.log

Building original tree took 0.7982666492462158 seconds

The masked_filtered-delim.iqtree.log file is empty.

What else do I need to modify in order for the workflow to give me bootstrap values?

I’m still constantly learning new things about how Nextstrain works so apologies if my understanding of something is off-base.

Thank you!

An update on this. I tried running this with -b 100 instead of -bb 1000 (regular bootstrap instead of ultrafast bootstrap), which seemed to work up until it was time to compute the consensus tree. Then I got:


Reading tree(s) file results/masked_filtered-delim.fasta.boottrees ...
100 tree(s) loaded (0 rooted and 100 unrooted)
Trees have different taxa sets

It seems to me like the ‘/’ characters in the sample names (Country/samplename/date), which are fine for the default ncov workflow, seem to be causing problems here because they are replaced with different DELIM-XXX text in different trees?

So thus far, I haven’t succeeded in modifying --tree-builder-args to enable either ultrafast bootstrap or regular bootstrap.

I feel like I might be missing something obvious about how to get my bootstrap values onto my tree. Is this not something people working on SARS-CoV-2 regularly do?

Hi @aamtemp,

I can’t speak much on how to include bootstrap values in the tree, but the ncov workflow does not support
the override-default-args flag in the config yet. There’s currently an open PR to add this functionality to the ncov workflow.