Fitness Landscapes and Adaptive Evolution
summary
Study of accessible mutational pathways uses ~20 empirical 5–9 locus datasets, Fourier spectra, and random-field models (House-of-Cards, NK, Rough Mount Fuji). Key claims: House-of-Cards yields expected accessible paths = 1 but probability of any path → 0 as L grows; constraining the antipode flips the threshold at ~log L / L (Hegarty & Martinsson). NK block model gives huge mean path counts but vanishing accessibility; neighborhood scheme shifts fluctuations. Adaptive walks: random length ~log L, greedy ~1.7 steps, greedy reaches higher final fitness (≈1 − 0.44/L).
AI-written summary · about
chapters (9)
- 0:00Start
- 1:39Fitness landscapes
- 3:08Mathematical setting
- 6:32Measures of epistasis
- 11:14A metsanalysis of empirical data sets
- 17:18Null model: House-of-cards
- 19:17Distribution of number of accessible paths from reversal genotype
- 29:03Path decomposition for the block model
- 31:24Mean number of paths is insensitive to genetic architecture
related talks
44mNetwork Propagation as a Tool for Deciphering Disease Mechanisms
48mTight Genes: Intro to Genetic Algorithms
46mTight Genes: Intro to Genetic Algorithms
49mIntroduction to Biological Network Analysis IV: Network Alignment and Querying
56mIntroduction to Cancer Bioinformatics III: Cancer Progression Networks
57mSolutions That Evolve: Building Self-Improving Systems with Genetic Algorithms
01h 06mIntroduction to Biological Network Analysis III: Identifying Network Modules
45mRandom Walks on Directed Graphs