![Page 1: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/1.jpg)
Algorithmic Insightsand the Theory of Evolution
Christos H. Papadimitriou
UC Berkeley
![Page 2: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/2.jpg)
The Algorithm as a Lens
• It started with Alan Turing, 60 years ago
• Algorithmic thinking as a novel and productive way for understanding and transforming the Sciences
• Mathematics, Statistical Physics, Quantum Physics, Economics and Social Sciences…
• This talk: Evolution
![Page 3: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/3.jpg)
Evolution Before Darwin
• Erasmus Darwin
![Page 4: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/4.jpg)
Before Darwin
• J.-B. Lamarck
![Page 5: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/5.jpg)
Before Darwin
• Charles Babbage
[Paraphrased]
“God created not species, but the Algorithm for creating species”
![Page 6: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/6.jpg)
Darwin, 1858
•Common Ancestry•Natural Selection
![Page 7: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/7.jpg)
The Origin of Species
• Possibly the world’s most masterfully compelling scientific argument
• The six editions 1859, 1860, 1861, 1866, 1869, 1872
![Page 8: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/8.jpg)
The Wallace-Darwin papers
![Page 9: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/9.jpg)
Brilliant argument, and yet many questions left unasked, e.g.:
• How does novelty arise?
• What is the role of sex?
![Page 10: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/10.jpg)
After Darwin
• A. Weismann
[Paraphrased]
“The mapping from genotype to phenotype is one-way”
![Page 11: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/11.jpg)
Genetics
• Gregor Mendel [1866]
• Number of citations
between 1866 and 1901:
3
![Page 12: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/12.jpg)
The crisis in Evolution1900 - 1920
•Mendelians vs. Darwinians
•Geneticists vs. Biometricists/Gradualists
![Page 13: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/13.jpg)
The “Modern Synthesis”1920 - 1950
Fisher – Wright - Haldane
![Page 14: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/14.jpg)
Big questions remaine.g.:
• How does novelty arise?
• What is the role of sex?
![Page 15: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/15.jpg)
Evolution and Algorithmic Insights
• “ How do you find a 3-billion long string in 3 billion years?”
L. G.Valiant
At the Wistar conference (1967), Schutzenberger asked virtually the same question
![Page 16: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/16.jpg)
Valiant’s Theory of Evolvability
• Which traits can evolve?
• Evolvability is a special case of statistical learning
![Page 17: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/17.jpg)
Evolution and CS Practice:Genetic Algorithms [ca. 1980s]
• To solve an optimization problem…
• …create a population of solutions/genotypes
• …who procreate through sex/genotype recombination…
• …with success proportional to their objective function value
• Eventually, some very good solutions are bound to arise in the soup…
![Page 18: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/18.jpg)
And in this Corner…Simulated Annealing
• Inspired by asexual reproduction
• Mutations are adopted with probability increasing with fitness/objective differential
• …(and decreasing with time)
![Page 19: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/19.jpg)
The Mystery of Sex Deepens
• Simulated annealing (asexual reproduction) works fine
• Genetic algorithms (sexual reproduction) don’t work
• In Nature, the opposite happens: Sex is successful and ubiquitous
![Page 20: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/20.jpg)
?
![Page 21: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/21.jpg)
A Radical Thought
• What if sex is a mediocre optimizer of fitness (= expectation of offspring)?
• What if sex optimizes something else?
• And what if this something else is its raison d’ être?
![Page 22: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/22.jpg)
Mixability!
• In [LPDF 2008] we establish through simulations that:
• Natural selection under asex optimizes fitness
• But under sex it optimizes mixability:
• The ability of alleles (gene variants) to perform well with a broad spectrum of other alleles
![Page 23: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/23.jpg)
Explaining Mixability
• Fitness landscape of a 2-gene organism
Rows: alleles of gene A
Columns: alleles of gene B
Entries: fitnessof the combination
![Page 24: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/24.jpg)
Explaining Mixability (cont)
• Asex will select the largest numbers
![Page 25: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/25.jpg)
Explaining Mixability (cont)
• But sex will select the rows and columns with the largest average
![Page 26: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/26.jpg)
Neutral Theory and Weak Selection
• Kimura 1970: Evolution proceeds not by leaps upwards, but mostly “horizontally,” through statistical drift
• Weak selection: the values in the fitness matrix are very close, say in [1 – ε, 1 + ε]
![Page 27: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/27.jpg)
Changing the subject:The experts problem
• Every day you must choose one of n experts
• The advice of expert i on day t results in a gain G[i, t] in [-1, 1]
• Challenge: Do as well as the best expert in retrospect
• Surprise: It can be done!
• [Hannan 1958, Cover 1980, Winnow, Boosting, no-regret learning, MWUA, …]
![Page 28: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/28.jpg)
Multiplicative weights update
• Initially, assign all experts same probability
• At each step, increase the probablity of each by (1 + ε G[I, t]) (and then normalize)
• Theorem: Does as well as the best expert
• MWUA solves: zero-sum games, linear programming, convex programming, network congestion,…
![Page 29: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/29.jpg)
Disbelief
Computer scientists find it hard to believe that such a crude technique solves all these sophisticated problems
“The eye to this day gives
me a cold shudder.”
[cf: Valiant on three billion bits and years]
![Page 30: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/30.jpg)
Theorem [CLPV 2012]: Under weak selection, evolution is a game
•the players are the genes
•the strategies are the alleles
•the common utility is the fitness of the organism (coordination game)
•the probabilities are the allele frequencies
•game is played through multiplicative updates
![Page 31: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/31.jpg)
There is more…
• Recall the update (1 + ε G[i, t])
• ε is the selection strength
• (1 + ε G[i, t]) is the allele’s mixability!
• Variance preservation: multiplicative updates is known to enhance entropy
• Two mysteries united…
• This is the role of sex in Evolution
![Page 32: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/32.jpg)
Pointer Dogs
![Page 33: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/33.jpg)
Pointer Dogs
C. H. Waddington
![Page 34: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/34.jpg)
Waddington’s Experiment (1952)
Generation 1
Temp: 20o C
![Page 35: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/35.jpg)
Waddington’s Experiment (1952)
Generation 2-4
Temp: 40o C~15% changedSelect and breed those
![Page 36: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/36.jpg)
Waddington’s Experiment (1952)
Generation 5
Temp: 40o C~60% changedSelect and breed those
![Page 37: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/37.jpg)
Waddington’s Experiment (1952)
Generation 6
Temp: 40o C~63% changedSelect and breed those
![Page 38: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/38.jpg)
Waddington’s Experiment (1952)
(…)Generation 20
Temp: 40o C~99% changed
![Page 39: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/39.jpg)
Surprise!
Generation 20
Temp: 20o C~25% stay changed!!
![Page 40: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/40.jpg)
Genetic Assimilation
• Adaptations to the environment become genetic!
![Page 41: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/41.jpg)
Is There a Genetic Explanation?
Function f ( x, h ) with these properties:
•Initially, Prob x ~ p[0] [f ( x, h = 0)] ≈ 0%
•Then Probp[0][f ( x, 1)] ≈ 15%
•After breeding Probp[1][f ( x, 1)] ≈ 60%
•Successive breedings, Probp[20][f ( x,1)] ≈ 99%
•Finally, Probp[20][f ( x, 0)] ≈ 25%
![Page 42: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/42.jpg)
A Genetic Explanation
• Suppose that “red head” is this Boolean function of 10 genes and “high temperature”
“red head” = “x1 + x2 + … + x10 + 3h ≥ 10”
• Suppose also that the genes are independent random variables, with pi initially half, say.
![Page 43: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/43.jpg)
A Genetic Explanation (cont.)
• In the beginning, no fly is red (the probability of being red is 2-n)
• With the help of h = 1, a few become red
• If you select them and breed them, ~60% will be red!
![Page 44: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/44.jpg)
Why 60%?
![Page 45: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/45.jpg)
A Genetic Explanation (cont.)
• Eventually, the population will be very biased towards xi = 1 (the pi’s are close to 1)
• And so, a few flies will have all xi = 1 for all i, and they will stay red when h becomes 0
![Page 46: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/46.jpg)
Generalize!
• Let B is any Boolean function
• n variables x1 x2 … xn (no h)
• Independent, with probabilities
p = (p1 p2 … pn)
• Satisfiability game: if B is satisfied, each variable gets 1, otherwise 1 - ε
• Repeated play by multiplicative weights
![Page 47: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/47.jpg)
Boolean functions (cont.)
Conjecture: This solves SATCan prove it for monotone functions (in poly time)
Can almost prove it in general(Joint work with Adi Livnat, Greg Valiant, Andrew Won)
![Page 48: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/48.jpg)
Interpretation
• If there is a Boolean combination of a modestly large number of genes that creates an unanticipated trait conferring even a small advantage, then this combination will be discovered and eventually fixed in the population.
• “With sex, all moderate-sized Boolean functions are evolvable.”
![Page 49: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/49.jpg)
Sooooo…
• The theory of life is deep and fascinating
• Insights of an algorithmic nature can help make progress
• Evolution is a coordination game between genes played via multiplicative updates
• Novel viewpoint that helps understand the central role of sex in Evolution
![Page 50: Algorithmic Insights and the Theory of Evolution](https://reader035.vdocuments.us/reader035/viewer/2022062323/568159c2550346895dc712d1/html5/thumbnails/50.jpg)
Thank You!