Simulated datasets

This directory contains simulated datasets used to evaluate Contrast-FEL, simulation settings, and analysis results in JSON format. Please contact Sergei Pond (spond@temple.edu) with questions.

Simulated data

Null simulations

Data generated under the strict null, i.e. no differential selection. A compressed archive null.tar.bz, 11MB, which contains several directories each with

  1. replicate a simulation description file. This is a JSON file generated by the HyPhy simulation script, which includes a lot of details about the model used to generate the data. Of particular interest is the "simulator.site.profile" field, which, for each site records the values of α and two β rates (one for each set of branches).
  2. tree.nwk the partitioned tree used to generate data.
  3. replicate.replicate.XX a data replicate.
  4. replicate.replicate.XX.FEL.json the result of a Contrast-FEL analysis on the corresponding replicate.

Omnibus simulations

Data generated using latin hypercube sampling to explore the parameter space. A compressed archive ombinbus.tar.bz, 150MB, which contains a single directory with

  1. sim.XXX.settings a simulation description file. This is a JSON file generated by the HyPhy simulation script, which includes a lot of details about the model used to generate the data. Of particular interest is the "simulator.site.profile" field, which, for each site records the values of α and two β rates (one for each set of branches).
  2. sim.XXX.nwk is the partitioned tree used to generate data.
  3. sim.XXX.settings.replicate.N a data replicate from a particular scenario
  4. sim.XXX.settings.replicate.N.FEL.json the result of a Contrast-FEL analysis on the corresponding replicate.

Omnibus simulations with 4 branch sets

Data generated using latin hypercube sampling to explore the parameter space with 4 branch. A compressed archive ombinbus-multi.tar.bz, 22MB, which contains a single directory with the same notation as above.