This README describes the data and code for the manuscript:

"Shifts in mutation bias promote mutators by altering the distribution of fitness effects"

by
Marwa Tuffaha (Western University, London, Canada, mtuffaha@uwo.ca)
Saranya Varakunan (Western University, London, Canada, svarakun@uwo.ca)
David Castellano (University of Arizona, Arizona, USA, dcastellano@arizona.edu)
Ryan Gutenkunst (University of Arizona, Arizona, USA, rgutenk@arizona.edu)
Lindi Wahl (Western University, London, Canada, lwahl@uwo.ca)

The file "C_ToPublish.c" contains the code used to do the simulations showing the results in the manuscript.
It depends on subroutines from Numerical Recipes in C (ANSI-C edition). 
The code from Numerical Recipes is proprietary and is not reproduced here.

The code generates 3 text files for each replicate

1- First file named "N??_K??_bias??_mu??_replicate#.txt"
For each generation (row), it exports:
ntotal (total number of individuals in the population), 
ntypes (number of genotypes), 
ftbar / ntotal (mean bias of the population), 
ftseq[bigtype] (bias of the most common type), 
wbar / ntotal (mean fitness of the population), 
mubar / ntotal (mean mutation rate of the population), 
mus[bigtype] (mutation rate of the most common type)

2- Second file named "dfe_N??_K??_bias??_mu??_replicate#.txt"
For each checkpoint at which we write DFEs for the most common type (dfetimes), it exports:
igen (Generation number)     
fben (beneficial mutations fraction)       
fdel (deleterious mutations fraction)  
s_pos (mean positive selection coefficient)     
s_neg (mean negative selection coefficient)    
ws[bigtype] (fitness of the most common type)     
min_ham_dist (minimum hamming distance between the most common type and the initial genotypes)

3- Third file named "ts_tv_dfe_N??_K??_bias??_mu??_replicate#.txt"
For each checkpoint at which we write DFEs for the most common type (dfetimes), it exports:
igen (Generation number)
ts_fben (beneficial transitions fraction)
ts_fdel (deleterious transitions fraction)
ts_s_pos (mean positive selection coefficient in transitions)
ts_s_neg (mean negative selection coefficient in transitions)
tv_fben (beneficial transversions fraction)
tv_fdel (deleterious transversions fraction)
tv_s_pos (mean positive selection coefficient in transversions)
tv_s_neg (mean negative selection coefficient in transversions)
ws[bigtype] (fitness of the most common type)


All data generated from simulations using the code mentioned above are stored in csv format.
Note that there are no new original data associated with this study; these data are provided for convenience but can all be reproduced exactly using the simulation code and mathematical techniques described in the manuscript.
For reference, all filenames give the appropriate figure number and panel from the manuscript.

Rows in each file as described below. For each data row, we provide the following in order:

Variable name, Mathematical symbol, Units, Allowable values (any relevant parameter values that distinguish that row, in parentheses)
In any field, "--" indicates not applicable.


Fig2: The mutational bias affects the numbers of transition/transversion fixations. More transitions fix through time than transversions when the organism in transition-biased.
Fig2a:
row 1: time, --, generation, non-negative integer
row 2: mean number of mutation fixations, --, unitless, positive real number
row 3: mean number of transition fixations, --, unitless, positive real number
row 4: mean number of transversion fixations, --, unitless, positive real number
Fig2b:
row 1: time, --, generation, non-negative integer
row 2: fraction of transition fixations, --, unitless, real number [0,1] (bias, beta=0.9)
row 3: fraction of transition fixations, --, unitless, real number [0,1] (bias, beta=0.8)
row 4: fraction of transition fixations, --, unitless, real number [0,1] (bias, beta=0.7)
row 5: fraction of transition fixations, --, unitless, real number [0,1] (bias, beta=0.6)
row 6: fraction of transition fixations, --, unitless, real number [0,1] (bias, beta=0.5)
row 7: fraction of transition fixations, --, unitless, real number [0,1] (bias, beta=0.4)
row 8: fraction of transition fixations, --, unitless, real number [0,1] (bias, beta=0.33)
row 9: fraction of transition fixations, --, unitless, real number [0,1] (bias, beta=0.3)
row 10: fraction of transition fixations, --, unitless, real number [0,1] (bias, beta=0.2)
row 11: fraction of transition fixations, --, unitless, real number [0,1] (bias, beta=0.1)


Fig3: Beneficial fractions and beneficial effect-sizes decline during the evolution of a transition-biased population, while deleterious fractions increase.
Fig3a:
row 1: time, --, generation, non-negative integer
row 2: bias-weighted beneficial fraction, fb_obs, unitless, real number [0,1]
row 3: transitions beneficial fraction, fb_Ts, unitless, real number [0,1]
row 4: transversions beneficial fraction, fb_Tv, unitless, real number [0,1]
Fig3b:
row 1: time, --, generation, non-negative integer
row 2: relative difference in beneficial fraction, --, unitless, real number
Fig3c:
row 1: time, --, generation, non-negative integer
row 2: mean bias-weighted positive selection coefficient, --, unitless, real number [0,1]
row 3: mean positive selection coefficient in transitions, --, unitless, real number [0,1]
row 4: mean positive selection coefficient in transversions, --, unitless, real number [0,1]
Fig3d:
row 1: time, --, generation, non-negative integer
row 2: relative difference in positive selection coefficient, --, unitless, real number
Fig3e:
row 1: time, --, generation, non-negative integer
row 2: bias-weighted deleterious fraction, fd_obs, unitless, real number [0,1], --, unitless, real number [0,1]
row 3: transitions deleterious fraction, fd_Ts, unitless, real number [0,1], --, unitless, real number [0,1]
row 4: transversions deleterious fraction, fd_Tv, unitless, real number [0,1], --, unitless, real number [0,1]
Fig3f:
row 1: time, --, generation, non-negative integer
row 2: relative difference in deleterious fraction, --, unitless, real number

Fig4: Bias-shifted mutators have higher invasion probabilities (dashed) than mutators without a bias shift.
Fig4a through Fig4d:
row 1: Mutation Rate Multiple, F, unitless, non-negative real number
row 2: Invasion Probability, --, unitless, real number [0,1] (with no bias shift or epistasis)
row 3: Invasion Probability, --, unitless, real number [0,1] (with bias shift but no epistasis)
row 4: Invasion Probability, --, unitless, real number [0,1] (with no bias shift but with epistasis)
row 5: Invasion Probability, --, unitless, real number [0,1] (with bias shift and epistasis)


Fig5: Mutator invasion probability increases when bias is reversed.
Fig5a and Fig5b:
row 1: Starting bias, beta, unitless, real number [0,1]
row 2: Invasion Probability, --, unitless, real number [0,1] (with no bias shift or epistasis)
row 3: Invasion Probability, --, unitless, real number [0,1] (with bias shift but no epistasis)
row 4: Invasion Probability, --, unitless, real number [0,1] (with no bias shift but with epistasis)
row 5: Invasion Probability, --, unitless, real number [0,1] (with bias shift and epistasis)


Fig6: In populations that have evolved with a transition frequency of β = 0.9, the invasion probability of mutators increases when the bias is reduced or reversed.
Fig6a:
row 1: Shifted bias, beta', unitless, real number [0,1]
row 2: Invasion Probability, --, unitless, real number [0,1] (mutators emerge at bias-weighted beneficial fraction = 30%, with epistasis)
row 3: Invasion Probability, --, unitless, real number [0,1] (mutators emerge at bias-weighted beneficial fraction = 30%, no epistasis)
row 4: Invasion Probability, --, unitless, real number [0,1] (mutators emerge at bias-weighted beneficial fraction = 15%, with epistasis)
row 5: Invasion Probability, --, unitless, real number [0,1] (mutators emerge at bias-weighted beneficial fraction = 15%, no epistasis)

Fig6b:
row 1: Shifted bias, beta', unitless, real number [0,1]
row 2: Invasion Probability, --, unitless, real number [0,1] (F=200)
row 3: Invasion Probability, --, unitless, real number [0,1] (F=50)
row 4: Invasion Probability, --, unitless, real number [0,1] (F=10)