What is the meaning of final ML optimization likelihood value in raxml phylogenetic analysis?

Question

[Kumar]

I have generated a phylogenetic tree using raxml by following command

raxmlHPC -s FcC_supermatrix.phy -n species_tree -M -m GTRGAMMA -f a -x $RANDOM -N 1000 -p 12345 -q FcC_supermatrix_partition.txt

I have obtained a phylogenetic tree with well supported bootstrap value and my final ML Optimization Likelihood value is -45810.705251. Here, I do not know how to interpret the ML optimization Likelihood value. Please, somebody help me to interpret the same. Is it a acceptable value for the given phylogenetic tree?

Answer 1

[Mensur Dlakic]

There is no need to interpret the final log likelihood value because it has no meaning on its own. It is more meaningful to look at bootstrap values.

Roughly speaking, the final likelihood is inversely proportional to the size of your alignment (both in terms of length and number of sequences), so I would guess from your likelihood value that you have a relatively short alignment. These values only come into play when you are comparing trees made from the same alignment but with different parameter sets or with different programs. In that case higher likelihood value generally means more reliable tree.

[Kumar]

Thank you @Mensur Dlakic