Skip to content

7. nodetraits

Jump to bottom
T. Latrille edited this page Oct 4, 2023 · 2 revisions

Test of diversifying selection for a quantitative trait

Data formatting

To run the analysis on your dataset and compute posterior probabilities, the executable nodetraits and readnodetraits from BayesCode require three files:

  • A phylogenetic tree in newick format, with branch lengths in number of substitutions per site (neutral markers)
  • A file containing the mean trait values for each species.
  • A file containing the variation within species for each trait and the genetic variation within species (neutral markers).

1. Phylogenetic tree

The phylogenetic tree must be in newick format, with branch lengths in number of substitutions per site (neutral markers).

2. Mean trait for each species

The file containing mean trait values for each species must be in a tab-delimited file with the following format:

TaxonName Body_mass Brain_mass
Panthera_tigris 12.26 5.676
Pithecia_pithecia 7.256 3.436
Colobus_angolensis 9.176 4.284
Saimiri_boliviensis 6.845 3.279

The columns are:

  • TaxonName: the name of the taxon matching the name in the alignment and the tree.
  • As many columns as traits, without spaces or special characters in the trait.
  • The values can be NaN to indicate that the trait is not available for that taxon.

3. Trait variation for each species

The file containing trait variation for each species must be in a tab-delimited file with the following format:

TaxonName Nucleotide_diversity Body_mass_variance Body_mass_heritability Brain_mass_variance Brain_mass_heritability
Pithecia_pithecia 0.0016 0.22871 0.2 0.00737 0.2
Colobus_angolensis 0.0017 0.00393 0.2 0.00416 0.2
Saimiri_boliviensis 0.0013 0.00022 0.2 0.00045 0.2
Pygathrix_nemaeus 0.0016 0.00347 0.2 0.00097 0.2
  • TaxonName: the name of the taxon matching the name in the alignment and the tree.
  • Nucleotide_diversity: the nucleotide diversity within species (neutral markers), cannot be NaN.
  • As many columns as traits, without spaces or special characters in the trait.
  • TraitName_variance: the phenotypic variance of the trait within species, can be NaN to indicate that the trait variance is not available for that taxon.
  • TraitName_heritability (optional): the heritability of the trait within species, between 0 and 1, cannot be NaN.
  • The columns with the suffix _variance and _heritability are repeated for each trait.
  • TraitName_heritability_lower (optional): the lower bound of the heritability of the trait within species, between 0 and 1, cannot be NaN.
  • TraitName_heritability_upper (optional): the upper bound of the heritability of the trait within species, between 0 and 1, cannot be NaN.
  • If the columns with the suffix _heritability_lower and _heritability_upper are present, the heritability is randomly drawn from a uniform distribution between the lower and upper bounds.
  • If the columns with the suffix _heritability is present, it is taken as is.

If the genetic variance (instead of phenotypic variance) is available for a trait, the heritability can be omitted and will automatically be set to 1.0.

Running nodetraits and readnodetraits

The file data/body_size/mammals.male.tsv contains the mean trait values for each species, the file data/body_size/mammals.male.var_trait.tsv contains the variation within species for each trait and the genetic variation within species (neutral markers), and the file data/body_size/mammals.male.tree contains the phylogenetic tree.\

nodetraits is run with the following command:

nodetraits -t data/body_size/mammals.male.tree --traitsfile data/body_size/mammals.male.tsv --until 2000 run_mammals_male

Then the chain run_mammals_male is used to compute the posterior distribution of the ratio of between species variation over within species variation with readnodetraits:

readnodetraits --burnin 1000 --var_within data/body_size/mammals.male.var_trait.tsv --output results_mammals_male.tsv run_mammals_male

The file data_empirical/chain_name.ratio.tsv contains the posterior mean of the ratio of between species variation over within species variation, the 95% and 99% credible interval, and the posterior probability that the ratio is greater than 1.

Maximum likelihood estimation (optional)

To obtain the ratio (without the posterior credible interval and probability) using maximum likelihood computation, the following python script can be used:

python3 utils/neutrality_index.py --tree data/body_size/mammals.male.tree --traitsfile data/body_size/mammals.male.tsv --var_within data/body_size/mammals.male.var_trait.tsv --output results_ML_mammals_male.tsv
Clone this wiki locally