Genet. Res., Camb. (1961) 2 pp. 127-14With 2 text-figuresPrinted in Gr - PDF document

Genet. Res., Camb. (1961) 2 pp. 127-14With 2 text-figuresPrinted in Great BritainNatura selectio a th proces o accumulatin genetiinformatio i adaptiv evolutionB M0T0 KIMURNational Institute of Genetics, Mishima, Japan(Received 3 October 1960INTRODUCTIOModer geneti studie hav show tha th instruction fo formin 12 MOTO KIMTJRo fundamenta importanc whic await th fruitfu activitie o statistician anothe applie mathematician collaboratin wit biologistsTH CONCEP O A SUBSTITUTIONA LOAA uni proces i adaptiv evolutio i th replacemen i a Mendelia populatioo on allel b anothe whic i bette fitted t a ne environment I wa pointeou b Haldan (1957 tha i thi i carrie ou b prematur deat o les fit individuals i ma cos a numbe o death equa t abou thirt time th populationumber I propose th ter substitutiona (o evolutional loa t expres thdecreas o populatio fitness (i th Darwinia sense i th proces o suc a gensubstitutio (Kimura 1960a b).Le u conside th simples situatio i whic th populatio consist o haploiorganisms suc a som fungi I suc a case eac gen exist i a singl dos i asomati cell Le x b th frequenc (relativ proportion o a gen A whic i ith proces o bein substitute fo it allel A' becaus o it selectiv advantagove A'. The th rat o chang i gen frequenc x i give bodxwher s ( 0 i th selectiv advantag measure i Malthusia parameter(Fisher 1930) i.e i term o it contributio t th geometri growth-rat o thpopulation an t i th timeSinc th populatio a a give momen contain th unfi genotyp A' i th proportio o 1 — x, th tota decreas i populatio fitness, als measure i Malthusiaparameters throughou th proces o substitutio io 1 1L = js(l-x)dt = j^i-x)-^ = J = -log.*O pwher p i th initia valu o x. Thu w havL = -\ogep. (1Thi i th expecte substitutiona loa fo a haplon i th substitutio proceeda th rat o on gen pe generation Th actua substitutiona loa ma bobtaine b summin th abov quantit ove al relevan gen loci eac weighteaccordin t th rat o substitutio pe locu pe generationL,= 2eL -Seloga�j (2wher e i th rat o substitutio pe locusTh situatio i muc mor complicate fo highe animal an plants i whiceac gen exist i doubl dos withi a somati cel (diploidy an gen interactio Genetic information in adaptive evolution12withi a locu (dominance become important I ca b show (cf Kimura 196 b)that i th selectiv advantage o th genotype A A an A A' ove A'A' ar s ansh respectively the th loa produce b substitutin A fo A' i1-h{s ^ 0 1 ^ h ^ 0(3"h+(l-2h)pwher p i th initia frequenc o A i th populatio an h represent th degreo dominanc o A ove A' i fitness. On salien featur o thi resul i tha L doeno depen directl o s, th magnitud o selectiv advantag involvedI Fig l(a an (b), value o L ar plotte fo variou value o h an p. I ma4 6 810-2 2 4 6 8IQ- 2 4 6 810-Fig lab see fro thi figure tha L increase a p decreases whil i decrease a hincreases I man th typica frequenc o 'recessive deleteriou gene i o thorde o 1% an i w assum tha thei dominanc i fitness i abou 2 a irecessiv lethal o th fruit-fl Drosophila melanogaster, L turn ou t b abou 59A i th cas o th haplonts th substitutiona loa i give bL- =(4 13MOTO KlMTJRIQ-5Fig 16Fig 1 (a)and(b) Graph sho-win th substitutiona loa La a functio o initia genfrequenc p an degre o dominanc h. L i th decreas i fitnes whic i expectei th gen substitutio proceed a th rat o on gen pe generationTH SUBSTITUTIONA LOA A A MEASUR O GAI I GENETIINFORMATIOI no propos t sho tha th rat o accumulatio o geneti informatiodenote b H i directl proportiona t th substitutiona load namelyH =l-44Le bit pe generation(5wher bi i a commonl use uni o th amoun o informatio equivalen tth informatio conten o choosin betwee a pai o alternatives sa 0 o 1wit equa probabilit (0-5) Th abov relatio ma b derive fro tw independen course o reasoning(1 I thos individual whic ar t b eliminate b natura selectio i thproces o progressiv evolution wer kep aliv an allowe t reproduc a thsam rat a th favoure individuals th populatio numbe woul becomeafte t generations Genetic information in adaptive evolution 13time it initia value Thi mean tha natura selectio allow a inciden t occuwit probabilit one which withou selection coul occu onl wit a probabilitoThu informatio gaine throug t generation amount tan therefor informatio gaine pe generatio i^ --0 bitsa wa t b shown(2 Conside a populatio o haploi organisms Le p b th initia frequenc oa advantageous gen A. Th probabilit tha th gen A i ultimatel establishei th populatio i 1 unde natura selection whil i i only i natura selectiower no workin an th fixation o gene wer lef t th actio o rando genetidrift Thu th amoun o informatio correspondin t thi gen substitutio iH = log2- = -log2p bitsO th othe hand w hav show tha fo one-gen substitutio£ -\ogep. (cf (1)ThereforH = -\oe9erL = ,-^-ESTIMATIO O TH INFORMATIO GAI I THACTUA PROCES O EVOLUTIOA show above th gai o geneti informatio i directl proportiona t thsubstitutiona load an th proble o estimatin th forme (H) i no reduce ttha o estimatin th latte (Le). However sinc evolutio i usuall a exceedinglslo proces i compariso t ou ordinar life-span i ma b ver difficul tdetermin Le fro direc observation Haldan (1949 ha shown base o paleontologica dat o Simpso (1944) tha th standar rat o evolutio i morphologica character i o th orde o one-tent o a darwin on darwi standin fo achang b a facto o e pe millio yearsI a attemp t deriv theoretically som fundamenta geneti parameter suca mutatio rat (/x an degre o dominanc (h) fro th standar rat o evolutioi th past I propose wha I calle th principl o minimu geneti loa (Kimura19606) a hypothesi tha i th proces o evolutio th geneti parameter ten tb adjuste suc tha th tota geneti loa i minimized I particularLT = 13 MOTO KIMURma b minimize i adaptiv evolution Her Lm stand fo th mutationa loawhic arise throug th eliminatio o deleteriou gene produce b recurrenmutatio (cf Crow 1958) Base o thi principle i wa demonstrate tha thspontaneou mutatio rat pe gamete £ /x, an th harmoni mea o th degreo dominanc o mutan gene i fitness h, ca b derive fro th rat o substitutio o gene an th tota amoun o hidde deleteriou effect pe gamete omutan genes(6wher E i th rat o substitutio o gene i horoteli evolutio (standar ratevolution cf Simpson 1944)E = S«an D i th tota amoun o geneti damag pe gamet expresse i letha equivalent (cf Morton Cro & Muller 1956) I w tak E =^J^ a approximatvalu suggeste b Haldan (1957) an D = 2 th on obtaine b Morton Cro& Mulle fro th stud o inbreedin i man w geJi = 0-020 an 2 M = 0-0581bot o whic agre fairl wel wit th correspondin observe value i thfruit-fl Drosophila (k abou 2% 2 abou 4%) Thi i remarkabl sinc thcalculatio i base o a simplifie assumptio tha evolutio ha proceede a aconstan rat ove a indefinitel lon time Th calculatio als supplies a thsam time th substitutiona an mutationa loadsLe = 0-206 Lm = 0-099I wil tak Le = 0- an Lm = 0- fo th presen purposeThen th rat o accumulatio o geneti informatio become approximatelH = 0-2 bit pe generationi w appl relatio (5) Similarly w ma calculat th amoun o informatio b eliminatin deleteriou mutan gene b usin th relatio H = Lmjloge 2Thi i a amoun whic exactl cancel ou th los o informatio b mutationTabl 1 i a balance-shee o geneti informatio i evolutionTabl 1 Balance sheet of genetic information in the process ofhorotelic evolution (bits/generation)Gen substitutio + 0-2Appearanc o deleteriou mutan gene —0-1 " ..Eliminatio o th deleteriou mutan gene +0-1 JTota +0-2 Genetic information in adaptive evolution 13W ar no i a positio t estimat th tota amoun o geneti informatiowhic ha bee accumulate sinc th beginnin o th Cambria epoch Prio tthi w kno ver littl abou th actua form o lif o th eart becaus o thscarcit o fossi records Throug th followin epoch ou knowledg o th majocours o evolutio i fairl good Befor w d this i ma b instructiv t se hoeffectivel a hig leve o improbabilit i geneti organizatio ca b generate bnatura selection Wit H = 0-2 bi pe generation th amoun o geneti infor accumulate ove 1,00 generation i 29 bits O th othe hand accord t Eddington th tota numbe o electron i th univers i f 136.2256, oapproximatel 2-3 x 1079. Thus th probabilit tha a randoml chose electroou o th univers happen t b th preassigne on i th reciproca o thi numbean th correspondin measur o improbabilit i abou 26 bits Thi mean tha1,00 generation o natura selectio ca achiev somethin mor improbabl thathis But fo th actua proces o organi evolution a duratio o 1,00i a ver shor tim indeedW d no kno ho ol lif o eart is thoug ther ar som reason t believtha i ha existe 2 billio years W d know however tha i th Cambriaepoch whic starte abou 50 millio year ago th eart wa alread inhabiteb organism suc a jellyfish anneli worms trilobites crustaceans etcI w assum the tha th geneti informatio ha bee gaine a th rat o 0-2bi pe generation th tota amoun o geneti informatio accumulate sinc thbeginnin o Cambria epoc i0-29x5xl08 „,,- = l-45xlO8/Gwher G i th harmoni mea o th duratio o on generatio i years Unfortunately fo organism whic d no exis excep a fossils i i impossibl tmeasur th exac lengt o thei generations Al w ca d i t infe the frocontemporar analogues Fo variou group o animals ther i som tendenc fosmalle an les differentiate member t matur mor quickl tha th largean well-differentiate ones Now i th histor o evolution i i know tha iwa alway th forme typ o anima whic ha succeede i leavin descendantsFurthermore G i expecte t b muc smalle tha th arithmeti averag o thlength o on generation Wit n reliabl estimate availabl a present I assumea a biologicall reasonabl guess tha G i o th orde o on yearW ma conclud then tha th tota amoun o geneti informatio whic habee accumulate sinc th beginnin o th Cambria epoc alon th lineagleadin t highe mammal ma b o th orde o on hundre millio bit (108 bits)Correspondin t thi increas i geneti information ther ha occurre atremendou improvemen o phenotypi organizatio whic i implie i th terevolutio i th usua sens o th wordSTORAG AN TRANSFORMATIO O GENETI INFORMATIOOwin t th recen developmen o bacteria an vira genetic an als o DNchemistry i ha becom increasingl clea tha DN (deoxyribonuclei acid 13 MOTO KIMUBmolecule formin chromosome ar th carrier o geneti information Fro thistandpoint a chromosom ma b considere a a linea sequenc o nucleotidpairs o whic fou kind ar discriminatedMulle (1958 estimate th tota numbe o nucleotid pair whic ma b preseni th chromosom se o ma a approximatel 4 x 109, b dividin th tota maso DN containe i a huma sper (ca. 4 x 10~1 gr. b th mas o on nucleotidpai (ca. 10~2 gr.) Thus wit fou kind o nucleotid pairs th maximu amouno geneti informatio tha ma b store i th haploi chromosom se o maamount tIog244x10* = 8xlO9 bitsan twic a muc fo th diploi set1- x 101 bitsThi i th maximu amoun o geneti informatio tha ma possibl b store ith nucleu o a fertilize huma egg i th fou kind o nucleotid pair ar equallefficientI i generall accepte tha th informatio i DN i transferred vi RN(ribonuclei acid molecules t proteins an if a som worker i thi field assum Cric et al., 1957) a sequenc o thre nucleotide determine on o twentpossibl amin acids th abov valu shoul b reduce b a facto oIog220/log243 x 0-72givin1-1 xlO1 bitso roughl 101 bit a th maximu amoun o geneti informatio tha migheffectivel b store i th diploi chromosom setHer th chromosom se ma b compare wit a electroni computer IIB 650 fo example ther ar 2,00 memor location an i ca stor 20,00digits o abou 6-6 x 104 bit o information A mor interestin objec fo compariso i th self-reproducin machin envisage b vo Neumann Accordin tKemen (1955) vo Neumann' machin consist o a basi bo o 8 x 40 squareplu a tai containin 150,00 squares Th basi bo ha a functio analogou tth soma whil th tai contain th instruction o th machin an i analogou th chromosom set Thi tai consist o 150,00 cell whic ar i eithe a 'on 'of ' state an therefor i ma stor1- x 105 bito 'geneti information'Thes comparison no onl sho th tremendou complexit o chromosomstructure bu als revea a indee amazin efficac o DN codes—efficac o a extent a pointe ou originall b Mulle (1935) tha al o th chromosome presen initiall i th fertilize egg fro whic th presen populatio oth worl (som tw thousand o millions develope woul occup a volum abouequa t tha o a ordinar aspiri tablet Genetic information in adaptive evolution 13Decipherin DN codes i.e learnin t rea th geneti language i a verfascinatin proble whic wa vigorousl attacke fo th firs tim onl a feyear ag (cf Yeas 1958 and withou a Rosett stone i woul b solve onl bstatistica treatment thoug n succes seem t hav bee obtaine s farW hav estimated i th previou section tha th tota amoun o genetiinformatio whic ha bee accumulate sinc th Cambria epoc i o th orde 108 bits O th othe hand a show above th maximu amoun o genetiinformatio tha migh b store i th diploi chromosom se o ma ma b oth orde o a leas 101 bits I th firs estimat (108 bits i correct th differencbetwee thes tw estimate mus b real eve i w admi tha ou Cambriaancestor ha alread accumulate a considerabl amoun o geneti informationI so I believ tha thi differenc ca b interprete i tw ways namely eitheth amoun o geneti informatio whic ha bee accumulate i a smal fractio wha ca actuall b store i th chromosom se or mor probably th DNcod itsel i highl redundant I a stimulatin paper t whic mor attentioshoul b pai b Wester geneticists Schmalhause (1958) a Russia geneticistpoint ou tha highe reliabilit o transmitte informatio ma b achieve bth repetitio o information suc a repetitio o equa gene (polygenes) 'repeatso gen complexe an i particula diploid o polyploidy Furthermore therma b repetitio o a certai kin o redundanc o informatio withi eac gen 'cistron'Recently Sueok (1960 ha mad a extensiv surve o th guanine-cytosin(G-C conten o deoxyribonuclei aci (DNA take fro variou organism rangin fro bacteri t man I ha bee foun tha fo vertebrate th averag contenlie withi th rang o 4 t 44% Fo variou specie o bacteria i varie ove amuc wide rang o 2 ~ 75% indicatin th marke divergenc i phylogenyO th othe hand th G- conten o DN molecule withi a organis ha arathe narro distributio wit 2c (twic th standar deviation o som 6 o les aroun it specifi mea valu p. I th cas o nativ DN take froth cal thymus th averag G- content p, i abou 40 an it heterogeneity2a i 9-6% whic ha bee th larges valu eve observedI ma no b difficul t se tha i th arrangemen o G- pair an A- pairwer entirel a random th proportion p, o G- pair betwee molecule withia organis shoul b distribute unimodall wit a binomia varianc o2 _ p(l-p)°P - —r~� (7wher b stand fo th numbe o bas pair (su o G- an A- pairs composina DN molecule I th actua case th numbe o bas pair pe molecul i noconstan an w shoul us it harmoni mea fo b. I th cas o cal thymuDNA th harmoni mea i roughl 104 an i mos case b shoul b a leassevera thousands Thus wit b = 104, th expecte heterogeneit (2ap) calculatefo p = 0- ~ 0- i0-00 2ap 0-01 13 MOTO KIMURi.e i lie betwee 0- an 1-0% I b i hal a large th heterogeneit wil becomabou 1- time a large an eve i 6 i I a large i wil onl doubl th abovvalue O th othe hand th observe heterogeneit i som 6 i vertebrate(Sueoka 1960) muc large tha tha expecte fro th binomia distributionThu th rati betwee th observe an th expecte heterogeneit i roughl 6i term o standar deviatio an 3 i term o varianceI assum tha thi discrepanc betwee th observe an th expecte i du trepetitio i th patter o bas arrangemen withi th DN molecule A simplanalysi o ou ordinar languag wil hel u greatl t clarif thi point I iknow that i Englis sentences th mos frequen lette i 'e' an thi i followeb ' t o ' a i frequency Thes thre letters togethe wit a spac betwee wordsmak u abou 40 (analogou t G- pair i 'geneti language') I extractefift lines eac wit sevent lette positions fro a pape o genetic an calculateth mea frequenc of e' ' t' ' a an spac pe lin an th standar deviatio oth frequenc betwee differen lines A show i th secon ro o Tabl 2 thrati betwee th observe an th expecte standar deviatio i abou 0-92.Her th expecte standar deviatio i calculate fro (7 usin p = 0-43 anb = 70 Simila calculation wer performe fo line take fro genetica paperwritte i French Germa an Russian Th result ar als liste i Tabl 2Tabl 2 Mean and standard deviation of the relative frequency of the sum of'e', 't', 'a' and space per line in samples of seventy letter positions. Thefigures in the last column denote the ratio between observed and expectedstandard deviations. For each language, fifty lines were extracted forcalculation.LanguagEnglisFrencGermaRussiaMea (p)0-430-450-380-29Standar deviatio (a)observe0-0540-0530-0540046Standar deviatio (ap)expecte00590-06100600053Rati0-920-870-86I ma b see fro thi tabl tha th rati betwee observe an expecteheterogeneit in e t a space conten pe lin i roughl 0-9 whic i ver differenfro wha w hav foun i th G- conten pe DN molecule. Suppos wduplicat eac lette fifty time s tha eac lin no consist o 5 x 7 o 3,50lette positions B thi duplication th observe heterogeneit doe no changbu th expecte heterogeneit ma b reduce t abou 1/7- becaus b shoul btake a 3500 Then th rati betwee th observe an expecte become abou6-4 whic i simila t th rati obtaine fo DNAGrantin tha ther ma b som thirty-si repetition i th arrangemen o* Thi valu i mainl du t th negativ correlatio betwee neighbourin letters Givinvalu 1 fo ' e' ' t' ' a an spac an valu 0 fo th remainin letters I obtaine correlatiocoefficient o —0-2 betwee tw adjacen letters —0-0 betwee tw neighbourin letteronc removed etc•j Se not a en o paper Genetic information in adaptive evolution 13bas pair i DN molecule th nex questio i ho suc repetition ca bvisualized Th analog wit ou ordinar languag ma hel u againLe u take a a example th followin sentenceI I SOFro thi sentenc w ca deriv variou form o lette arrangement b duplicatin eac lette twice Amon them th followin thre ar especiall significantwhic I wil tentativel cal letter wor an sentenc repetition respectively(i Lette repetition IIT IIS SSO(ii Wor repetition ITI ISI SOS(iii Sentenc repetition I I S I I SThes thre form are a such indistinguishabl wit respec t th rati betweeth observe an th expecte 'heterogeneity' However b splittin eac othes int piece o certai lengt an studyin thei heterogeneity w wil fin thathe behav quit differently Suppos w spli eac int tw piece o equa lengthI th cas o lette repetition th observe varianc wil becom double becauseac piec contain onl on hal o independen letter o th origina sentence Oth othe hand i th cas o sentenc repetition th wil remai th sambecaus eac piec contain th ful sentence Th situatio i intermediat i thcas o wor repetitio an th varianc become 5/3 Returnin t th probleo heterogeneit i G- content Sueok (1959 foun tha b splittin cal thymuDN molecule b ultrasoni vibratio int piece o abou one-tent i size thheterogeneit varianc increase ver littl (orde o 10 i any) Sinc w havalread assume tha ther ma b thirty-si repetition o letter i DNA thiresul ma b interprete a showin tha repetitio i th DN molecul mus bmor nea t th typ o sentenc repetitio tha tha o wor repetition Assumin104 bas pair pe DNA the eac sentenc consist o roughl 30 lettersTh abov hypothesi ma b teste b splittin DN molecule int muchsmalle piece consistin o les tha 30 bas pair an b seein i th observean th expecte heterogeneit agre wit eac otherI shoul b note her tha th repetitio ma no b exac i th actua DNmolecule rathe a eac repetitio th ' word ma b slightl modifie fro on tth other lik variation i musicA an rate throug th proces o individua developmen (ontogeny) thgeneti informatio i finall transforme int phenotypi information wit itvariou aspect i morphology physiolog an behaviour admittedl a largamoun o redundanc bein involve amon them Then ho larg i th phenotypi informatio o highe mammals o specificall tha o man Perhap thmor pertinen questio t as her i ho muc mor phenotypi informatio icontaine i highe animal o ma a compare t thei Cambria ancestors I sense, th informatio conten shoul no b counte i term o atomi omolecula configurations bu shoul b don i term o th three-dimensiona structur plu chemica data a pointe ou b Elsasse (1958) H 13 MOTO KIMURsuggest that sinc th informatio conten o huma specie pertainin t grosanatom alon coul hardl b diagramme o a plan are o 1 m2 i whic thsmalles uni o discriminatio i 1 mm2, an sinc gros anatom ca onl b amoderat fractio o th informatio conten o th organism th informatio conten o th huma organis mus b a leas o th orde o 107 bit or mor probably108 bits Elsasse state tha eve a figure o 109 bit woul hardl appea fantasticHowever sinc th phenotypi informatio i transforme geneti information thforme canno b large tha th latter whic w hav estimate a bein o thorde o 108 bits Th correspondenc th geneti an phenotypi informatio turn ou t b quit clos considerin that whil ne geneti informatioca onl b gaine throug natura selectio actin o genotypes thi actio imediate b th phenotype whic ar determine b th genotypes A morreliabl estimat wil b supplie i th futur b anatomist o chemist wh wil acces t a prope statistica methodologyI m opinio th creativ rol o natura selection whic i stil no infrequentloverlooke b evolutionists ma mos convincingl b brough t ligh b calculatin it powe o accumulatin geneti informatio an considerin th phenotypicomplexit a it product Lerne (1959 state tha th meanin o natura selectioa a creativ proces ma b well illustrate b quotin Michelangelo' concep ocreation 'Th sculptor' han ca onl brea th spel t fre th figures slumberini th stone. Indeed an elaborat wor o ar mus contai a larg amoun oinformationSUMMAR1 I th cours o evolution complicate organism hav descende fro mucsimple ones Sinc th instruction t for a organis ar containe i thnucleu o it fertilize egg thi mean tha th geneti constitutio ha becomcorrespondingl mor comple i evolution I w expres thi complexit i termo it improbability definin th amoun o geneti informatio a th negativlogarith o it probabilit o occurrenc b chance w ma sa tha geneti informatio i increase i th cours o progressiv evolution guide b natura selectioo rando mutations2 I wa demonstrate tha th rat o accumulatio o geneti informatio iadaptiv evolutio i directl proportiona t th substitutiona load i.e th de o Darwinia fitness brough abou b substitutin fo on gen it allelifor whic i mor fitted t a ne environment Th rat o accumulatio o genetiinformatio i give bH = -—e— x l-44ie ('bits'/generation)wher Le i th substitutiona loa measure i 'Malthusia parameters'3 Usin Le = 0-199 a valu obtaine fro th applicatio o th 'principl ominimu geneti load (cf Kimura 19606) w geH = 0-2 bit/generation Genetic information in adaptive evolution 13I wa estimate tha th tota amoun o geneti informatio accumulate sincth beginnin o th Cambria epoc (50 millio years ma b o th orde o108 bits i evolutio ha proceede a th standar rateSinc th geneti informatio i transforme int phenotypi informatio iontogeny thi figur (108 bits mus represen th amoun o informatio whiccorrespond t th improve organizatio o highe animal a compare t theiancestor 50 millio year back4 Problem involve i storag an transformatio o geneti informatio thuacquire wer discusse an i wa pointe ou tha th redundanc o informatioi th for o repetitio i linea sequenc o nucleotid pair withi a gen mapla a importan rol i th storag o geneti informationREFERENCECHICK F H C GRIFFITH J S & ORGEL L E (1957) Code withou commas Proc. nat.Acad. Sci., Wash., 43 416-421CROW J F (1958) Som possibilitie fo measurin selectio intensitie i man Hum. Biol.30 1-13ELSASSEE W M (1958) The Physical Foundation of Biology. London Pergamo PressFISHER R A (1930) The Genetical Theory of Natural Selection. Oxford Clarendo PressHALDAJNE J B S (1949) Suggestion a t quantitativ measuremen o rate o evolutionEvolution, 3 51-56HALDANE J B S (1957) Th cos o natura selection J. Oenet. 55 511-524KEMBNY J G (1955) Ma viewe a a machine Sci. Amer. 192 58-67KIMXTRA M (1960a) Geneti loa o a populatio an it significanc i evolution (Japaneswit Englis summary. Jap. J. Genet. 35 7-33KIMURA M (19606) Optimu mutatio rat an degre o dominanc a determine b thprincipl o minimu geneti load J. Genet. 57 21-34LERNER I M (1959) Th concep o natura selection A centennia view Proc. Amer. Phil.Soc. 103 173-182MORTON N E. CROW J F & MULLER H J (1956) A estimat o th mutationa damagi ma fro dat o consanguineou marriages Proc. nat. Acad. Sci., Wash., 42 855-863MUIXER H J (1929) Th metho o evolution Sci. Mon., N.Y. 29 481-505MULLER H J (1935. Out of the Night. Ne York Vanguar PressMULLER H J (1958) Evolutio b mutation Bull. Amer. math. Soc. 64 137-160SCHMALHATTSEN I I (1958) Contro an regulatio i evolution (Russia wit Englissummary. Bull. Soc. Nat. Moscow, 63 93—121SIMPSON G G (1944) Tempo and Mode in Evolution. Ne York Columbi Universit PressStrEOKA N (1959) A statistica analysi o deoxyribonuclei aci distributio i densitgradien centrifugation Proc. nat. Acad. Sci., Wash., 45 1480—1490SUEOKA N (1960) Som geneti an evolutionar consideration o th bas compositio odeoxyribonuclei acids (I press.YCAS M (1958) Th protei text Symposium on Information Theory in Biology, pp 70-102London Pergamo PressNOT ADDE I PROOAfte thi pape ha bee sen t press I ha th privileg o seein a preprin oa pape b J Josse A D Kaise an Kornberg wh successfull determineth neares neighbou sequenc o nucleotide i DN take fro variou organisms Fro thei Tabl VI I calculate th correlatio betwee tw adjacennucleotid pair i cal thymu DNA givin valu 1 fo a G- pai an 0 fo a 14 MOTO KIMURA— pair Th correlatio coefficien obtaine wa abou — 0-09 a valu no drasticall differen fro th on obtaine fo Englis sentences Simila calculationfo bacteria an bacteriophag DNA' (Table VII an IX gav correlatio coefficient o a mos a fe pe cen (eithe positiv o negative) Cf Josse Kaise &Kornber (1960) Enzymati synthesi o deoxyribonuclei acid VIII Frequencieo neares neighbo bas sequence i deoxyribonuclei acid Jour. biol. Ghent, (ipress)