/BUSTED-SRV

The effect of synonymous rate variation on BUSTED

Primary LanguageHyPhyArtistic License 2.0Artistic-2.0

BUSTED[S]

A repo with data, scripts and information associated with the forthcoming BUSTED[S] publication.

HBL

includes BUSTED-SRV-sim.bf the script used to generate simulation replicates.

batch_script.sh an example of the slurm batch script used to run simulation replicates through BUSTED an BUSTED[S]

Data

to find the simulation data from the paper please visit: http://data.hyphy.org/web/busteds/

includes original examples of data and results jsons from BUSTED[S]

paper_LFs includes likelihood function batch files (.LF.bf) used to generate the 16 sequence and 31 sequence simulation replicates in the paper as well as newick trees for both when omega = 1.

BUSTED-SRV sim parameters - Inputs for sims.csv shows the naming convention and parameter values used for each simulation of 100 replicants. Due to size limitations on github actual sequences will be hosted on silverback (TBD).

How to use BUSTED-SRV-sim.bf and analyze the data

Note: These small tutorials assume a wokring installation of HyPhy and a basic working knowledge of the HyPhy command line interface, R, and bash.

Create your own simulated data set

  1. Generate or use pre-existing Likelihood Function file (.LF.bf)
    • pre-existing can be found in the paper_LFs folder
  2. Run HYPHYMP BUSTED-SRV-sim.bf You will be prompted for
    • .LF.bf
    • Synonymous rate categories and weights
    • Non-synonymous rate categories and weights
    • number of replicants
    • the path and name you wish to save them to (eg: /home/sadie/simulations/Sim_A will give you files Sim_A_replicant_1 etc.)

Use the pre-exisiting parameters

if you would like to use the paramters from the paper to regenerate the simulated data sets do the following:

  1. Download the BUSTED-SRV sim parameters - Inputs for sims.csv file, the paper_LFs and gen_batch_cmd_list.R
  2. using gen_batch_cmd_list.R change the relevant file paths
  3. Run R gen_batch_cmd_list.R to get a .txt file of bash commands
  4. Run the bash commands to generate all of the replicates (may not want to run all at once)

Analyzing the data

  • There are several possible ways to run the analysis on a large number of files but the basic command used for this paper is:

HYPHYMP busted --alignment {input} --srv Yes --grid-size 2000 --starting-points 10 --output {output}

  • you can use bash, snakemake, or your favorite tool to loop over all the replicates in a folder and analyze them

  • We've provided a snakemake example in Snakefile and an older bash example in batch_script.sh