中子通量与病毒突变率的滞后互相关:把’天时改疫’从巧合变成可检验假设

Written by

in

Language

Evidence level: B

把”星象致病”翻译成可计算的统计问题:中子计数序列领先病毒突变率序列几个月,还是两者根本不同步?

正文

易学”天时”最硬的版本是”岁运影响疫变”:天体节奏不只是外环境背景,而是直接改写病毒进化。上一篇文章(gcr-neutron-flux-virus-mutation-rate)梳理了这条候选物理通道——太阳活动调制银河宇宙线(GCR),地面中子监测站把通量变化记录成计数序列。争议的核心不是”通道是否存在”,而是”证据姿势”:拿疫情爆发年份去对太阳极小期,是选择偏差的重灾区(2008-2009 极小期通量同样高,却没有新病毒登场,PMID 35194566 的反证就是证据)。正确的检验姿势是时序对时序:取中子监测站的日/月通量序列,与同一时段呼吸道病毒的系统发育 root-to-tip 突变速率序列做滞后互相关——通量升高是否在数月后跟随突变率上升?

这条设计有两个技术支点,都已成熟。第一个是 root-to-tip(RTT)回归:把序列的采样时间做横轴、遗传分歧度做纵轴,回归斜率就是分子钟速率。它是分子钟推断的标准工具,从艾滋病病毒整合时间估计(PMID 36632480)到 SARS-CoV-2 起源时间回溯(PMID 32687861)都用它;2024 年的 Clockor2 把 RTT 从命令行升级成可交互的网页工具,支持全局与局部分子钟的同时拟合(PMID 38366939)。第二个是滞后互相关/分布滞后模型:流行病学早已习惯”暴露今天、发病数天后”的错位,流感与空气污染的月度滞后关系可以精确到 6 个月滞后窗(PMID 37726705),南京 13 万 SARI 病例的时间序列研究也证实高污染暴露的滞后效应”出现更早、持续更久”(PMID 35130812)。

但把两个成熟工具接起来之前,必须先承认三个”模型内敌人”,否则任何显著相关都可能是伪影。

第一,突变速率不是一个稳定的常数。 对 50 种 RNA 病毒的定量系统发育分析早就表明:速率存在显著的谱系间变异,严格分子钟常被拒绝(PMID 11821909)。更麻烦的是时间依赖速率现象(TDRP):速率估计随观测时间尺度系统性衰减,幂律斜率约 -0.65,其机制已被建模为”最快进化位点饱和 + 纯化选择”的组合(PMID 34478645)。这意味着”突变率序列”本身携带强烈的内在趋势——把它跟同样有趋势的中子通量序列做相关,即使物理上毫无关联,也可能因为共享低频漂移而得到虚假的正相关。去趋势(detrending)或对趋势建模,是第一步,不是可选项。

第二,突变率的方差主要由病毒自己的发动机解释。 SARS-CoV-2 的进化率约 9.90×10⁻⁴ 替换/位点/年(PMID 32687861);奥密克戎登场后突变速率近乎翻倍,从 2021 年 12 月前的每年约 23 个替换增至之后约 45 个(PMID 36768588)——这个跳变对应的是免疫逃逸选择与人群传播动力学,而不是任何天体指数。因此检验的正确表述是”增量信息量”:把宿主免疫(疫苗覆盖率、既往感染率)与气象(温度、湿度)作为协变量塞进模型后,中子通量的边际贡献还剩多少。若中子通量的滞后系数在控制协变量后塌缩到零,则”天时改疫”的直接通道不成立,效应必须经由气候/宿主免疫等间接通路实现——这与五运六气需要跟气候变量耦合使用而不是孤立断言,恰好是同一个结论。

第三,样本量与被试数。 RTT 速率估计需要足够的序列-时间跨度支撑;对单个疫情爆发做”几个月对几个月”的互相关,自由度小得可怜。可行的设计是把时间轴拉长到 10-20 年、跨多病毒(流感、RSV、冠状病毒、呼吸道合胞病毒),用每年/每季度的滑动窗口估计”窗口内速率”,再与同窗口的中子通量序列对齐。这本质上是把”岁运对应”从事件史(对不上就尴尬)换成连续谱(永远有数据可比对)——统计上更诚实,也更容易出可证伪的结论。

对易学的意义:这项检验把”天时”还原成一条可测量的物理通道(太阳活动→日球层调制→宇宙线通量→病毒诱变)并给出明确的证伪条件。若中子通量在控制一切已知协变量后仍有显著滞后增量,则”岁运影响疫变”获得近代天文-生物学双锚点;若不显著(当前证据倾向于此),则”天时”的效应应归入气候与宿主免疫等间接通路——两种结局都有科学价值,且都不否定五运六气作为”气候-健康耦合框架”的实用意义,只是把它的作用层级从”直接改写病毒基因组”降为”通过环境间接调制”。

证据与局限

已验证: 中子监测站计数是宇宙线通量的可靠代理,2019 深极小期通量确实偏高(PMID 35194566);RTT 回归是成熟的分子钟推断工具(PMID 36632480, 38366939);RNA 病毒进化率约 10⁻³ 替换/位点/年且存在显著谱系间变异(PMID 11821909, 32687861);流行病学滞后模型(GAM/DLNM)能刻画暴露-发病的数月错位(PMID 37726705, 35130812)。

部分验证: 流感血凝素突变存在约 11 年周期、太阳黑子极值与流感大流行 OR=3.87(PMID 27136236);SARS-CoV-2 突变速率随免疫选择出现翻倍跳变(PMID 36768588)——但该跳变与天体周期指数不同步,提示速率方差主要由宿主免疫与传播动力学主导。

待验证: 中子通量序列与呼吸道病毒 root-to-tip 突变速率的滞后互相关是否存在;去趋势与协变量控制后中子通量的边际增量信息量;高海拔 vs 低海拔(宇宙线通量数倍差异)同病毒监测点的解耦度对比。

相关阅读

证据等级: B(RTT 方法学、RNA 病毒速率变异、滞后模型、TDRP 机制均有 PubMed 直接支撑;”中子通量→病毒突变率”的直接滞后关联检验缺失,属科学类比/间接证据链)
参考文献: PMID 35194566, PMID 36632480, PMID 38366939, PMID 11821909, PMID 32687861, PMID 34478645, PMID 37726705, PMID 35130812, PMID 36768588


免责说明:本文属传统知识与科学文献的整理、对照与研究设计讨论,仅供理解与参考,不构成医疗建议,不能替代专业诊断与治疗。文中涉及的效应量、频率区间与暴露结论多为研究报告层面的推断,个体情况差异很大;如有身体不适或健康疑虑,请就医并由专业医师判断。

Lag Cross-Correlation Between Neutron Flux and Viral Mutation Rates: Turning \”Changing the Weather to Stop Epidemics\” from Coincidence into a Testable Hypothesis

Evidence level: B

Translate the concept of “astral causation” into a calculable statistical problem: Do neutron count sequences lead viral mutation rate sequences by several months, or are they fundamentally out of sync?

Main Text

The most robust formulation of Yijing “heavenly timing” is the “solar cycle influences epidemic change”: celestial rhythms are not merely an external environmental backdrop but directly rewrite viral evolution. The previous article (gcr-neutron-flux-virus-mutation-rate) outlined this candidate physical channel—solar activity modulates Galactic Cosmic Rays (GCR), and ground-based neutron monitors record flux variations as count sequences. The core controversy is not whether the “channel exists,” but the “evidentiary posture”: matching epidemic outbreak years against solar minima is a prime example of selection bias (the counter-evidence from PMID 35194566 shows high flux during the 2008-2009 minimum yet no new viral emergence). The correct testing posture is time-series against time-series: take daily/monthly flux sequences from neutron monitors and perform a lagged cross-correlation with the systemic phylogenetic root-to-tip mutation rate sequences of respiratory viruses during the same period—does flux increase follow by months with a rise in mutation rates?

This design rests on two technical pillars, both mature. The first is root-to-tip (RTT) regression: sampling time serves as the x-axis, genetic divergence as the y-axis, and the regression slope represents the molecular clock rate. It is the standard tool for molecular clock inference, used from HIV integration time estimation (PMID 36632480) to SARS-CoV-2 origin tracing (PMID 32687861); the 2024 Clockor2 tool upgrades RTT from a command-line interface to an interactive web platform, supporting simultaneous global and local clock fitting (PMID 38366939). The second is lagged cross-correlation/distribution lag models: epidemiology has long been accustomed to the misalignment of “exposure today, onset days later,” with monthly lag relationships between influenza and air pollution precise to a 6-month lag window (PMID 37726705); a time-series study of 130,000 SARI cases in Nanjing also confirms that high-pollution exposure effects “appear earlier and persist longer” (PMID 35130812).

However, before connecting two mature tools, one must acknowledge three “internal model enemies”; otherwise, any significant correlation may be an artifact.

First, mutation rates are not stable constants. Quantitative phylogenetic analyses of 50 RNA viruses have long demonstrated significant lineage-specific rate variation, where strict molecular clocks are frequently rejected (PMID 11821909). More problematic is the Time-Dependent Rate Phenomenon (TDRP): rate estimates systematically decay with observed time scales, with a power-law slope of approximately -0.65, a mechanism modeled as a combination of “fastest-evolving site saturation + purifying selection” (PMID 34478645). This implies the “mutation rate sequence” itself carries strong intrinsic trends—correlating it with a similarly trending neutron flux sequence may yield false positive correlations due to shared low-frequency drift even if physically unrelated. Detrending or modeling the trend is a prerequisite, not an option.

Second, the variance of mutation rates is primarily explained by the virus’s own engine. SARS-CoV-2 evolves at approximately 9.90×10⁻⁴ substitutions/site/year (PMID 32687861); after Omicron emerged, the rate nearly doubled, increasing from ~23 substitutions/year before December 2021 to ~45 (PMID 36768588)—this jump corresponds to immune escape selection and population transmission dynamics, not any celestial index. Therefore, the correct formulation of the test is “incremental information”: after inserting host immunity (vaccine coverage, prior infection rates) and meteorology (temperature, humidity) as covariates into the model, what remains of the marginal contribution of neutron flux? If the lag coefficient for neutron flux collapses to zero after controlling for covariates, then the direct channel of “heavenly timing alters epidemics” is invalid, and effects must operate via indirect pathways such as climate or host immunity—this aligns with the conclusion that Wu Yun Liu Qi (Five Movements Six Qi) requires coupling with climate variables rather than isolated assertion.

Third, sample size and subjects. RTT rate estimation requires sufficient sequence-time span support; performing a “months against months” cross-correlation on a single epidemic outbreak yields negligible degrees of freedom. A feasible design extends the time axis to 10-20 years, spanning multiple viruses (influenza, RSV, coronaviruses, respiratory syncytial virus), estimating “windowed rates” using annual/quarterly sliding windows, then aligning with neutron flux sequences within the same window. Essentially, this transforms “solar cycle correspondence” from an event history (awkward if mismatched) to a continuous spectrum (always data-comparable)—statistically more honest and yielding easier-to-falsify conclusions.

Significance for Yijing Studies: This test reduces “heavenly timing” to a measurable physical channel (solar activity → heliosphere modulation → cosmic ray flux → viral mutagenesis) and provides explicit falsification conditions. If neutron flux retains significant lagged incremental information after controlling all known covariates, “solar cycle influences epidemic change” gains a modern astronomical-biological dual anchor; if not significant (current evidence leans this way), the effect of “heavenly timing” should be attributed to indirect pathways via climate and host immunity—both outcomes hold scientific value and do not negate the practical utility of Wu Yun Liu Qi as a “climate-health coupling framework,” merely lowering its operational level from “direct rewriting of viral genomes” to “indirect modulation via environment.”

Evidence and Limitations

Verified: Neutron monitor counts are reliable proxies for cosmic ray flux; 2019 deep solar minimum flux was indeed elevated (PMID 35194566); RTT regression is a mature molecular clock inference tool (PMID 36632480, 38366939); RNA virus evolution rates are ~10⁻³ substitutions/site/year with significant lineage-specific variation (PMID 11821909, 32687861); epidemiological lag models (GAM/DLNM) can characterize multi-month misalignments between exposure and onset (PMID 37726705, 35130812).

Partially Verified: Influenza hemagglutinin mutations exhibit an ~11-year cycle; solar sunspot extrema correlate with influenza pandemics (OR=3.87) (PMID 27136236); SARS-CoV-2 mutation rates show a doubling jump due to immune selection (PMID 36768588)—yet this jump is out of sync with celestial cycle indices, suggesting rate variance is dominated by host immunity and transmission dynamics.

Pending Verification: Whether lagged cross-correlation exists between neutron flux sequences and respiratory virus root-to-tip mutation rates; the marginal incremental information of neutron flux after detrending and covariate control; decoupling comparisons between high-altitude and low-altitude sites (cosmic ray flux differs by orders of magnitude) versus virus monitoring points.

Related Reading

Evidence Level: B (RTT methodology, RNA virus rate variation, lag models, and TDRP mechanisms have direct PubMed support; “neutron flux → viral mutation rate” direct lagged correlation testing is missing, belonging to scientific analogy/indirect evidence chain)
References: PMID 35194566, PMID 36632480, PMID 38366939, PMID 11821909, PMID 32687861, PMID 34478645, PMID 37726705, PMID 35130812, PMID 36768588


Disclaimer: This article compiles traditional knowledge and scientific literature for comparison and research-design discussion only. It is not medical advice and cannot replace professional diagnosis or treatment. Effect sizes, frequency ranges and exposure conclusions cited here are largely inferences at the research-report level, and individual circumstances vary widely. If you have any discomfort or health concerns, please consult a physician.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *