Trouble with options --max_reticulations and --brlen

[max-hill]

Thank you for writing and sharing such a great program!

I am using NetRAX to analyze some viral datasets (~10 taxa, ~150kbp), but I have had some difficulty getting the program to run with certain values of the options --max_reticulations and --brlen. Hopefully the errors I'm having are just user error, but I've gotten stuck so I thought I would reach out and ask.

When I run NetRAX with the following command, the program runs correctly, and returns a network with 6 reticulations:

mpiexec ./netrax --msa ../../../analysis/netrax/experiment-max1/dataset.fasta --model ../../../analysis/netrax/experiment-max1/partition.txt --output ../../../analysis/netrax/experiment-max1/netrax-output-linked-max10-reticulations.txt --average_displayed_tree_variant --start_network ../../../analysis/netrax/experiment-max1/experiment-max1.treefile --max_reticulations 10 --seed 42 --brlen linked

But when I reduce the number of reticulations to a number less than 6, the software runs for a little while and then terminates with a runtime error while searching for an arc removal move. For example, the following command, in which I have reduced the number of reticulations from 10 to 1, produces such an error (see log1.txt):

mpiexec ./netrax --msa ../../../analysis/netrax/experiment-max1/dataset.fasta --model ../../../analysis/netrax/experiment-max1/partition.txt --output ../../../analysis/netrax/experiment-max1/netrax-output-linked-max1-reticulation.txt --average_displayed_tree_variant --start_network ../../../analysis/netrax/experiment-max1/experiment-max1.treefile --max_reticulations 1 --seed 42 --brlen linked

Similarly, when I try running the original command using the unlinked branch lengths option, the program experience a seg fault. For example:

~/virus-project/scripts/NetRAX/bin$ mpiexec ./netrax --msa ../../../analysis/netrax/experiment-max1/dataset.fasta --model ../../../analysis/netrax/experiment-max1/partition.txt --output ../../../analysis/netrax/experiment-max1/netrax-output-unlinked-max10-reticulations.txt --average_displayed_tree_variant --start_network ../../../analysis/netrax/experiment-max1/experiment-max1.treefile --max_reticulations 10 --seed 42 --brlen unlinked
optimizing model, reticulation probs, and branch lengths (normal mode)...
BIC score after model optimization: 522531.061
--------------------------------------------------------------------------
Primary job  terminated normally, but 1 process returned
a non-zero exit code. Per user-direction, the job has been aborted.
--------------------------------------------------------------------------
--------------------------------------------------------------------------
mpiexec noticed that process rank 0 with PID 0 on node mutalisk exited on signal 11 (Segmentation fault).
--------------------------------------------------------------------------

Do you have any insight about what might be going on here?

Thanks again, Max

[lutteropp]

Hi Max,

Thank you for your interest in using NetRAX! :-)

It looks like you found some bugs in the implementation. We did not use the max_reticulation parameter or the unlinked parameter in our experiments. Looks like the code needs more test coverage! Also for unlinked branch lengths, I vaguely mean to remember that this option was not fully implemented and still had a TODO in its code.

I'll take a look and fix the issues. If possible, can you share the input files with me? It would help a lot with recreating the bug situation faster.

Best regards, Sarah

[max-hill]

Okay, thank you so much! The input files I used for this were these: input-files.zip

Let me know if you need anything else.