search

abacus-gene / paml

PAML is a program package for model fitting and phylogenetic tree reconstruction using DNA and protein sequence data. Please report only **technical issues** on this repository (e.g., compiling, programs abort or do not run at all, etc.). Problems with input data and general questions should be posted at https://groups.google.com/g/pamlsoftware?pli
GNU General Public License v3.0
103 stars 19 forks source link

mcmctree terminated with errors "resetting lnL" when usedata =2 #44

Closed yemingyue222 closed 5 months ago

yemingyue222 commented 5 months ago

Hello, I meet a problem when using approximate likelihood calculation during the MCMC. Please help me.

My phylogenetic tree is over 100 taxa, I want to shorten the estimation of divergence time, so I use approximate likelihood calculation. The first step "usedata = 3" is smooth, but the next step "usedata = 2"in nohup.out file, e.g., 910275614542723227185095639040.000000 = 910275614542725760459886034944.000000? Resetting lnL Similar sentences appear five times. Please check my compressed files including 4 input files (input.tre input.tre.zip , input.phy input.phy.zip , mcmctree2.ctl mcmctree2.ctl.zip , in.BV) and 4 output files output files.zip (nohup.out, mcmc.out, mcmc.txt, SeedUsed), I would sincerely appreciate your advice.

Look forward to your reply! Best regards,

Mingyue YE

yemingyue222 commented 5 months ago

The compressed file of In.BV is more than 25 Mb and failed to upload. :(

sabifo4 commented 5 months ago

Hi there!

Thanks for your message, but please note that there is a PAML discussion group where this type of questions should be posted as then all users can benefit from the answers given by the PAML community :)

I have checked your input files and I can see that you do not have a constrain in your root age in your tree file. Instead, you just added option RootAge in the control file specifying an upper bound of max. 1.0 (time unit), which is currently deprecated. I just wanted to check that this was indeed the constrain you wanted to apply for your root age. Option finetune is deprecated, so there is no need to include it in the control file.

As I cannot see the in.BV file, I cannot really troubleshoot what may have gone wrong there. Regarding the format of this file, you should make sure that there are as many blocks with branch lengths, gradient, and Hessian in the in.BV file as alignment blocks you have in your alignment file (i.e., 353 according to variable ndata in your control file). They also should be in the same order as the alignment blocks in your alignment file (first block with branch lengths, gradient, and Hessian in the in.BV file should correspond to the first alignment block in your alignment file, etc.). Can you confirm that the in.BV follows this format?

Please post a message in the PAML discussion group and attach a compressed file with the in.BV file, control file, alignment file, and input file that you are using to continue troubleshooting this issue -- please let us know whether there was an issue with the in.BV file you were using or whether the issue still persists after making sure the format is correct.

I will close this issue here as it is not a technical problem related to MCMCtree. Nevertheless, I shall reopen this issue if a technical problem is later found :)