Skip to content

Whether Being Advantageous Matters Is Decided by the Size of the Population ── A mutation with a selection coefficient of 0.01 fixes 19.8013 times as often as a neutral one at N=500, but only 1.2012 times as often at N=10 ── what decides is not s but Ns ── and at s=0 the probability is exactly the neutral value ── [Paper 478]

Sep 2026 · Zenodo (CERN European Organization for Nuclear Research)
Evolution and Genetic Dynamics

Abstract

An advantageous mutation is said to spread. This paper shows that what decides whether the advantage matters is not its size but its product with the population size──where Ns is small, being advantageous does no good. No new theorem or law is claimed. Scope of this paper (scope note): No new theorem or law is claimed──Kimura’s fixation probability, genetic drift and the nearly neutral theory are all standard. No evolutionary history is argued──particular traits, lineages and dates are not the subject. The diffusion approximation is used──non-overlapping generations, constant population size and weak selection are assumed; real populations often violate these. Effective size is not distinguished──N is treated as the effective population size, which can differ greatly from the census count. Linkage and recombination are excluded──hitchhiking, background selection and linkage disequilibrium are not the subject. Relation to earlier papers: Paper 477 showed that adding to a non-minimal element does nothing──there too, it was a pairing and not a quantity alone that acted. Paper 300 showed that common root and distinct root can be decided──“selection is acting” and “Ns is large” are distinct roots by that criterion. Paper 446 showed that a floor moves with the demand──there too, a quantity alone was undetermined without its condition. What is added is giving the fixation probability at four population sizes, giving the ratio to neutral as 1.2012 and 19.8013, confirming that the ratio rises monotonically with N, showing that Ns=0.02 lies within five per cent of neutral, confirming the exact neutral value at s=0, and placing the separator at s versus Ns. First, a mutation with s=0.01 saturates at a fixation probability of 0.019801 even at N=10000 (Section 2). Second, a neutral one is 1/(2N), falling without bound as N grows (Section 2). Third, this is the core of the paper. The ratio to neutral is 1.2012 at N=10 and 19.8013 at N=500 (Section 3). Fourth, the ratio rises monotonically with N, approaching neutral as Ns shrinks (Section 3). Fifth, at s=0 the fixation probability is exactly 1/(2N) (Section 3). Sixth, the separator is whether the advantage acts or Ns acts (Section 4). An advantageous mutation is said to spread──advantage means being selected, apparently the very definition. But at first only one individual carries it: chance can remove it whatever its advantage, and how much chance matters is set by how small the population is. Advantage acts only in large populations──a mutation with s=0.01 fixes with probability 0.019801 at N=1000 and the same at N=10000: once the population is large enough, its fate no longer depends on the size. That value is close to 2s──even a one per cent advantage spreads only once in about fifty tries. The neutral case meanwhile falls from 0.050000 to 0.000050, a thousandfold. So in absolute terms the small population fixes more often, not because of advantage but because chance spreads things there. The ratio runs from 1.2012 to 19.8013──at N=10 (Ns=0.10) it is only 1.2012, barely distinguishable from neutral: being advantageous is doing almost no work. At N=500 (Ns=5.00) it is 19.8013, with the same s=0.01──the advantage unchanged, only the population differing. The ratio rises monotonically with N, with 1.2012<4.0342<19.8013 holding at the intermediate points too: the smaller Ns, the closer to neutral. At Ns=0.02 it is 1.0302, within five per cent of neutral, and in the limit it coincides. And at s=0 the probability is exactly 1/(2N)──a boundary and a check on the expression. A deleterious mutation can fix in a small population too: 0.8215 times neutral at s=-0.01 and N=10, against 7.738x10^-172 at N=10000. The same disadvantage, a different fate by size. One thing separates them──does the advantage act, or does Ns. If the advantage acts, s tells all and one need look only at the trait. If Ns acts, nothing follows without the population, the same s giving 1.2 or 19.8. They are distinct roots by the criterion of Paper 300, one a property of a trait and one of a trait paired with a population. It is the same shape as Paper 477, where what acted was not how much was added but where: in both, a quantity alone does not settle the answer. It overlaps Paper 446 too, where “how many times the floor” was undetermined without the recovery: here “an advantageous mutation” is undetermined without the population size. So harmful mutations can accumulate in small populations, because selection cannot sort them out──this is the content of the nearly neutral theory, and it matters for populations near extinction. To be honest──this rests on the diffusion approximation, assuming non-overlapping generations, constant size and weak selection, which real populations often violate. 19.8013 belongs to the approximation; that what acts is Ns and not s holds broadly across this family of models. On the making of this work: The ideas and content of this work stem from the author's own considerations. Assistance from an AI (a large language model) was used for structuring, English translation, and checking the algebra. Any remaining errors or misinterpretations are solely the author's. Feedback and corrections are sincerely appreciated. Keywords: fixation probability, genetic drift, population size, natural selection, nearly neutral theory. ----- 有利な変異は広がる、と言われる。本稿が示すのは、有利さが効くかどうかを決めているのは、有利さの大きさではなく集団の大きさとの積であることである──Ns が小さければ、有利であることが役に立たない。新しい定理も法則も主張しない。 本稿の射程(射程注記):新しい定理も法則も主張しない──木村の固定確率の式、遺伝的浮動、ほぼ中立説はすべて既知である。進化の歴史を論じない──特定の形質、系統、年代は主題ではない。拡散近似を用いる──世代が重ならず、集団サイズが一定で、選択が弱いという仮定に立つ。実際の集団はこれらを満たさないことが多い。有効集団サイズを区別しない──N を有効集団サイズとして扱う。実際の個体数とは大きく違いうる。連鎖と組換えを扱わない──ヒッチハイキング、背景選択、連鎖不平衡は主題ではない。既刊との関係:論文477 は最小でないものを足しても効かないことを示した──そこでも、量そのものではなく組が効いていた。論文300 は同根か別根かが判定できることを示した──「選択が働く」と「Ns が大きい」は、その基準で別根である。論文446 は下限が要求で動くことを示した──そこでも、単独の量が条件なしには決まらなかった。加えたのは、四つの集団サイズで固定確率を出したこと、中立との比を 1.2012 倍と 19.8013 倍と出したこと、比が N について単調に増えることを確かめたこと、Ns=0.02 で中立から 5 パーセント以内に入ることを示したこと、s=0 で厳密に中立値になることを確かめたこと、分離子を「s か Ns か」に置いたことである。 第一に、s=0.01 の変異は、N=10000 でも固定確率が 0.019801 で頭打ちになる(第2節)。 第二に、中立の場合は 1/(2N) で、N とともにいくらでも小さくなる(第2節)。 第三に、これが本稿の芯である。中立との比は N=10 で 1.2012 倍、N=500 で 19.8013 倍である(第3節)。 第四に、比は N について単調に増え、Ns が小さいほど中立に近づく(第3節)。 第五に、s=0 では固定確率が厳密に 1/(2N) になる(第3節)。 第六に、分離子は「有利さが効くか、Ns が効くか」である(第4節)。 有利な変異は広がる、と言われる──有利なら選ばれる、というのは選択の定義そのものに見える。だが最初は一個体しか持っていない。有利でも偶然で消えうるし、偶然の大きさは集団の小ささで決まる。有利さは、大集団でしか効かない──s=0.01 の変異の固定確率は N=1000 でも N=10000 でも 0.019801 で頭打ちになる。集団が十分大きければ、有利な変異の運命は集団サイズに依らない。その値は 2s にほぼ等しい──1 パーセント有利な変異でも、約 50.5 回に 1 回しか広がらない。一方、中立の場合は 0.050000 から 0.000050 へ 1000 分の 1 になる。だから絶対値では小集団のほうが固定確率が高いが、それは有利だからではなく、偶然で広がりやすいだけである。比を見ると、1.2012 倍から 19.8013 倍まで動く──N=10(Ns=0.10)では 1.2012 倍にしかならず、中立とほとんど区別がつかない。有利であることが、ほぼ役に立っていない。 N=500(Ns=5.00)では 19.8013 倍になる。同じ s=0.01 で、有利さは変わらず集団だけが変わった。比は N について単調に増え、1.2012<4.0342<19.8013 と途中でも順序が保たれる。 Ns が小さいほど中立に近い。 Ns=0.02 では比が 1.0302 と中立から 5 パーセント以内に入り、極限では中立に一致する。そして s=0 では固定確率が厳密に 1/(2N) になる。境目であり、式の確認でもある。不利な変異でも小集団では固定しうる──s=-0.01、N=10 で中立の 0.8215 倍、N=10000 では 7.738x10^-172 倍で実質ゼロ。同じ不利さが、集団の大きさで運命を変える。分けているものは一つ──有利さが効くか、Ns が効くか。有利さが効くなら s を見れば分かり、1 パーセント有利ならどこでも 1 パーセント有利で、形質だけを見ればよい。 Ns が効くなら集団を知らなければ分からず、同じ s が 1.2 倍にも 19.8 倍にもなる。論文300 の基準で別根であり、一方は形質の量、一方は形質と集団の組の量である。論文477 と同じ形で、そこでは足した量ではなくどこへ足したかが効いた。どちらも、単独の量では答えが決まらない。論文446 とも重なり、そこでは「下限の何倍か」が回収率を言わなければ決まらなかった。ここでは「有利な変異」が集団サイズを言わなければ決まらない。だから小集団では有害な変異も溜まりうる。選択が振り分けきれないからで、これがほぼ中立説の中身であり、絶滅の危機にある集団で問題になる。正直に言えば──拡散近似に立っている。世代が重ならず、集団サイズが一定で、選択が弱いという仮定であり、実際の集団はこれらを満たさないことが多い。19.8013 という値は近似のものだが、効くのが s ではなく Ns であるという結論は、この族の模型で広く成り立つ。 作成にあたって:本稿の着想と内容は、著者自身の考察に基づくものです。文章の構成整理や英訳、数式の確認には AI(大規模言語モデル)の助力を得ました。最終的な内容の解釈や誤りがあれば、それらはすべて著者の責に帰します。お気づきの点があれば、ご教示いただければ幸いです。 キーワード:固定確率、遺伝的浮動、集団の大きさ、自然選択、ほぼ中立説。

View source

Similar papers

#small language model Dataset Open access Oct 2026

Socratic guiding questions in synthetic arithmetic data: matched LoRA runs (revision v2)

Supporting data, adapters, predictions and code for the article *Low-Cost LoRA Fine-Tuning of Small Language Models for Multi-Step Arithmetic Reasoning* by Jake O'Grady, Asena Isik Gürhan, Chee Fong Ting and Effirul Ramlan (University of Galway). We generated 20,000 GSM8K-derived arithmetic problems with step-by-step s...

O'Grady, Jake, Gürhan, Asena Isik, Chee, Fong Ting et al. · 465 citations
#computer vision Open access Jun 2016

Software Development in Startup Companies: The Greenfield Startup Model

The results are packaged in the Greenfield Startup Model (GSM), which explains the priority of startups to release the product as quickly as possible, and the need to shorten time-to-market, by speeding up the development through low-precision engineering activities.

Carmine Giardino, Nicolò Paternoster, M. Unterkalmsteiner et al. · 178 citations · ⚡14
#computer vision Open access Oct 2016

Software Startups - A Research Agenda

Software startup companies develop innovative, software-intensive products within limited timeframes and with few resources, searching for sustainable and scalable business models.

M. Unterkalmsteiner, P. Abrahamsson, Xiaofeng Wang et al. · 157 citations · ⚡17
#machine learning Review Open access Oct 2016

“Failures” to be celebrated: an analysis of major pivots of software startups

This study conducts a case survey study based on the secondary data of the major pivots happened in 49 software startups, and demonstrates that customer need pivot is the most common among all pivot types.

Sohaib Shahid Bajwa, Xiaofeng Wang, Anh Nguyen-Duc et al. · 127 citations · ⚡15
#computer vision Review Open access May 2015

A survey study on major technical barriers affecting the decision to adopt cloud services

The comparison of adopter and non-adopter sample reveals three potential adoption inhibitor, security, data privacy, and portability, which underlines the importance of the technical and security perspectives for research investigating the adoption of technology.

Nattakarn Phaphoom, Xiaofeng Wang, S. Samuel et al. · 111 citations · ⚡8
#computer vision Open access Feb 2018

Lean Internal Startups for Software Product Innovation in Large Companies: Enablers and Inhibitors

This study investigates how Lean internal startup facilitates software product innovation in large companies and identifies its enablers and inhibitors, and shows the potential of the method-in-action framework to investigate the Lean startup approach in non-startup context.

Henry Edison, Nina M. Smørsgård, Xiaofeng Wang et al. · 78 citations · ⚡6

Related blog posts

MIT News · Artificial Intelligence Oct 15, 2026

AI model decodes the language cells use to communicate

A new computational approach reveals how cells communicate to assemble into tissues. This could accelerate stem cell engineering, improve disease models, and inform therapies.

Microsoft Research Blog Oct 12, 2026

Simulating student learning for better AI tutoring

StudentSim is a research system for evaluating and improving AI tutoring tools before they are tested with learners. It models how individuals may engage with tutors, giving feedback to improve tutor performance across subjects like chess, writing, and mathematics. The post Simulating student learning for better AI tutoring appeared first on Microsoft Research.

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.