Slim model of TEs

Open Tasks

• [X] Haploid model
• add cli options to diploid model
• store parameters in JSON file for default values?
• [X] add code for batch runs with GNU parallel

Parameters and Variables

• ‘N’: Number of individuals
• ‘L’: Genome length (mutations are stackable, so we can use length 1 for efficiency)
• ‘simTime’: simulation time in generations
• ‘teInitialCount’: initial number of TEs in each individual (heterozygous in the diploid model)
• ‘teJumpP’: probability of a TE to jump to a random genomic position
• ‘teDeathRate’: probability of lethal transposition

Run simulations from command line

slim -d N=10 -d L=1 -d teJumpP=0.01 -d teDeathRate=0.0005 -d simTime=100 TE_haploid_WIAS.slim

Batch simulations from commandline

Simulation replicates can be run in parallel on 3 or any number of cores with GNU parallel.

seq 10 | parallel -j 3 -n0 "slim -d N=5000 -d L=1 -d teJumpP=0.01 -d teDeathRate=0.0005 -d simTime=2000 TE_haploid_WIAS.slim"

Future considerations from biology

Recombination and genome structure

How does recombination affect the evolutionary equilibrium of TEs? Choice of the genome size L? What about the cost of replicating a larger genome? Optimization of genome size L, from position-independent deleterious effect of each TE? Realistic genomic density Or alternatively more comples up to continuous distribution of fitness effects for TEs, dependent of genomic context and region. Do TE jumps satisfy the infinite sites model of mutation, or, how common does a TE jump into another TE? What about variable recombination landscapes, as repetitive sequence motifs can favor certain recombination breakpoints.

Deactivation of TEs

Deactivation of TE due to loss of function (mutational load or loss of function of TE coding region), can be modelled at a steady rate (see SLiM manual section 14.12). Each TE can have a life-cycle, from active (autonnomous) to semi-active (non-autonomous) to dead (pseudo-gene). How do non-autonomous TEs transpose? Is it just an increase probability of gene transfer? The progression of a single TE through these stages is irreversible, and would lead to accumulation of dead TE sequences as determined by the fitness cost of replication.

Alternatively, TE activation can be of active, epigenetic nature from the ‘host’. In such case, how are the TEs deactivated? In large bursts? At potentially disruptive positions? Epigenetically, inactivated TEs could reactivate again as well, potentially mediated by environmental conditions or epigenetic reprogramming.