FeaturesAndDifferences

• Potential use cases of MimicrEE2
  • Features of MimicrEE2
  • Differences between MimicrEE1 and MimicrEE2

Potential use cases of MimicrEE2

Following some examples of potential use cases for MimicrEE2. Whenever applicable we also provide links to publications that already performed the tasks and in brackets the software that was used.

Improving the design of E&R studies

• evaluate the suitability of different experimental designs (e.g using ROC curves https://www.ncbi.nlm.nih.gov/pubmed/24214537 ; MimicrEE1); E&R studies are a major investment, both in terms of money and time (e.g. sequencing cost, selection for many generations). Optimizing the experimental design will thus be invaluable to increase the return on this investment (e.g.: to more accurately identify the loci contributing to adaptation).
• visualize the simulation results with Manhattan plots; MHP provide an intuitive overview of the effect of different experimental designs and of expectations under different adaptive scenarios (e.g. https://www.ncbi.nlm.nih.gov/pubmed/24214537 MimicrEE1)
• could novel selection regimes where, for example, the strength of selection is gradually increased during E&R studies enhance the power to identify the targets of selection (work in progress; MimicrEE2).

Evaluating the suitability of statistical approaches for E&R studies

• evaluate the suitability of different test statistics for identifying the targets of selection; Many different test statistics have been proposed for identify selected loci (e.g. cmh-test, diff-stat, association statistic, Fst, Gaussian process test https://www.ncbi.nlm.nih.gov/pubmed/25269380 ; MimicrEE1; https://www.ncbi.nlm.nih.gov/pubmed/25614471 ; MimicrEE1). Simulations will help to identify the most suitable test statistics.
• validate novel approaches, for example to reconstruct haplotype blocks from E&R data (https://academic.oup.com/mbe/article/34/1/174/2420816 MimicrEE1)

test hypothesis that may explain genomic patterns observed in E&R studies

• could diminishing returns epistasis lead to stagnating response to selection (e.g. https://academic.oup.com/mbe/article-abstract/35/1/180/4563780; SLiM)?
• could selection of a rare haplotype lead to an elevated long range linkage disequilibrium (e.g. https://academic.oup.com/mbe/article/31/2/364/997936 MimicrEE1) ?
• could directional selection lead to larger haplotype blocks than stabilizing selection?
• could polygenic adaptation explain a pattern of genetic redundancy, where beneficial alleles solely respond in a subset of the replicate populations (https://www.biorxiv.org/content/early/2018/05/28/332122 ; MimicrEE2)
• could recessive deleterious alleles be responsible for a stagnating response to selection (as for example observed in http://onlinelibrary.wiley.com/doi/10.1111/j.1365-294X.2012.05673.x/full)?
• could deleterious alleles linked to migrant haplotypes explain an observed pattern of low frequency of migrant alleles in evolving population (work in progress; MimicrEE2)

Note MimicrEE1 offers a subset of the functionality provided by MimicrEE2, hence all simulations performed with MimicrEE1 may also be performed with MimcrEE2

Features of MimicrEE2

MimicrEE2 supports the following features

• forward simulations with millions of SNPs located on different chromosomes
• large populations (several thousand) may be used
• a variable recombination rates that changes along chromosomes
• arbitrary haplotypes of the base population
• de novo mutations
• neutral simulations
• an arbitrary number of replicates
• classically selected loci (i.e. loci having a constant selection coefficient and heterozygous effect, this allows to simulate additivity, recessivity, dominance, overdominance, etc)
• thousands of selected SNPs having an arbitrary effect size
• arbitrary complex epistasis for pairs of SNPs
• arbitrary complex migration regimes
• diploid and haploid organism
• clonal evolution
• quantitative traits, with loci having arbitrary effect sizes and dominance; either the heritability or the environmental variance may be provided
• truncating selection for a quantitative trait and temporarily variable truncating selection
• stabilizing selection for a quantitative trait
• disruptive selection for a quantitative trait
• diminishing returns epistasis for a quantitative trait
• directional selection for a quantitative trait
• arbitrary complex fitness functions for a quantitative trait (i.e. mappings of phenotype to fitness)
• adaptation to a moving optimum (e.g. the fitness functions changing during the simulations)
• variable population sizes
• output of the allele frequencies at the requested generations (using the sync format which allows to proceed directly with PoPoolation2)
• output of the haplotypes of evolved populations at requested generations
• output the genotypic value, phenotypic value, and fitness of the evolved populations at the requested generations
• conversion of haplotypes into fasta files, which enables simulating reads with third party tools (e.g. ART https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3278762/)
• scripts that allow to plot simulation results as Manhattan plots
• scripts that help to run MimicrEE2 on a Appache Spark cluster
• a detailed manual and many walkthroughs
• thorough validation
• Scripts that allow to generate several of the required input files
• Script that allow converting haplotypes from Arlequin/Fastsimcoal to the Mimicree2 format

Differences between MimicrEE1 and MimicrEE2

Novel features in MimicrEE2 (as compared to MimicrEE1)

• as the major difference: in addition to classical selected loci (MimicrEE1), a quantitative trait model is implemented in MimicrEE2
• de novo mutations
• truncating selection, stabilizing selection, directional selection, diminishing returns epistasis, arbitrary complex fitness functions
• adaptation to a moving optimum
• arbitrary complex epistasis
• arbitrary complex migration regimes
• variable population size
• the recombination rate may be provided in multiple distinct units (recombination fraction, mean of Poisson distribution, cM/Mb)
• either the haplotypes or the allele frequencies or both may be saved as output
• the genotypic values, phenotypic values and the fitness of individuals may be reported
• conversion of haplotypes to fasta files, which enables simulations of reads (e.g. using ART https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3278762/)
• simulations of haploids
• simulations of clonal evolution
• conversion of arp files to MimicrEE2 haplotypes