'bucky.pl' in the TICR.

[nilofar alaie]

Hi everyone 

I run 'bucky.pl' in the TICR. 

It runs without any warning or error message. But I don't have .BUCKy.tar in my output directory and I just have .mbsum.tar.gz and .CFs.csv. 

and the CFS.csv does not look like the example file provided in the TICR examples. 

It should be 

# columns 1-4 should have these headers: taxon1 taxon2 taxon3 taxon4
# columns 5-7 should contain the CFs of quartets: 12|34, 13|24 and 14|23.

but it has two extra columns (_lo, _hi) for each CFs of quarter:

taxon1,taxon2,taxon3,taxon4,CF12_34,CF12_34_lo,CF12_34_hi,CF13_24,CF13_24_lo,CF13_24_hi,CF14_23,CF14_23_lo,CF14_23_hi,ngenes

Would you please help me to run the bucky for doing the TICR test?

The best

Niloo

[Cécile Ané]

The "lo" and "hi" columns give the lower and upper ends of the 95% credibility intervals. Each estimated concordance factor has its own credibility interval, which is why there are 3 "low" and 3 "high" values on each row: one interval for each of the 3 resolutions of a given four-taxon set.

If your .CFs.csv file has the results, then you should be all set.

[nilofar alaie]

Thanks for your respons. 

And I am wondering shall I remove the extra columns (''low'' and ''high'') for running the next step in TICR pipeline? 

I have 47 taxa and I should have at the end  178365  quartets in Bucky result but I don't know why it's just 177775   I used the slurm pipeline (http://crsl4.github.io/PhyloNetworks.jl/latest/man/ticr_howtogetQuartetCFs/) for running bucky, do you have any idea about this? 

and when I want to run the r script of getting the beach-length, getTreeBranchLengths.r, it gives me this error message: 

root for file names: example/CFtable

outgroup: last taxon

        tree was read. 47 taxa.

        listed the 44500 quartets associated with edges in tree.

Warning: it looks like there are missing (or extra) quartets in

        example/CFtable.CFs.csv.

        Read 177775 rows, expected 178365 quartets.

Error in `$<-.data.frame`(`*tmp*`, taxon1, value = character(0)) :

  replacement has 0 rows, data has 177775

Calls: $<- -> $<-.data.frame

Execution halted

do you think shall I remove the extra columns in bucky result? 

With the best 

Niloo

[Cécile Ané]

- I don't know why you have fewer 4-taxon sets than expected. Perhaps someone else on this list has an idea. If you have a file that logged the screen output, or some other log file (from slurm or your job scheduler), you might search this log file for an error message to see if some quartets failed to run (and which ones). You might be able to re-run them, if your job scheduler facilitates this.

- I suggest that you use the goodness of fit test in package QuartetNetworkGoodnessFit instead of TICR. First, the quartet goodness-of-fit test (in function quartetGoFtest!) can be done on a tree or a network more generally. Second, the quartet GoF test fixes a theoretical issue that is affecting the TICR test. You can follow the section on testing candidate networks to apply the test on your candidate tree. It shows how to get branch lengths on that tree (so no need to run the function getTreeBranchLengths.r, that gave you an error). This package will be okay with the extra columns that contain the low & high values of credibility intervals. :smile:

[nilofar alaie]

Thanks a lot for your answer. 

I still don't know, why I get the  fewer 4-taxon sets than expected, but I will see the log files as you said to see if there is any error message 

I can manage to run getTreeBranchLengths.r, just by modifying the cfs table and removing the extra (hi, low) columns.

Then I want to run the TICR.r script using the CFs table and tree with the branch length. 

 

But when I run the TICR.r, I get this error message: 

Error in `[.data.frame`(read.csv(buckyCF.filename), , tmp) :

  undefined columns selected

Calls: [ -> [.data.frame

Execution halted

I also try to run this script with the tree.tre and quartetCF.csv files in 'ticr/' directory, but I receive the same error.

Do you have any idea about this error? 

The best and thanks for your help 

Niloo