Heterologous Trained-Immunity Priming by Rabies Vaccination as a Candidate Prophylactic Countermeasure Against Filoviral Lethality: A Falsifiable Computational Hypothesis

Filoviral hemorrhagic fevers, Ebola virus disease (EVD) and Marburg virus disease (MVD), remain among the most lethal infections known, with outbreak case fatality ratios (CFRs) that can exceed 80% and no licensed pan filovirus prophylactic that is simultaneously inexpensive, thermostable adjacent, and deployable at primary care level in the low resource settings where filoviral outbreaks concentrate. This paper advances, as an explicitly labeled and falsifiable hypothesis rather than an empirical finding, the proposition that Human Diploid Cell Vaccine (HDCV) rabies immunization, optionally augmented with a beta glucan adjuvant, could confer clinically meaningful, non sterilizing attenuation of filoviral disease severity through the mechanism of trained immunity: durable epigenetic reprogramming of innate myeloid and natural killer (NK) effectors that partially overrides the VP35/VP24 mediated interferon antagonism central to filoviral immune evasion. We formalize this hypothesis as an extended Susceptible, Exposed, Infectious, Recovered, Deceased (SEIRD) compartmental model with a cross protection parameter epsilon and a lethality attenuation parameter delta, derive its reproduction number and eradication threshold algebra in closed form, and implement a fully reproducible, unit tested, seeded (NumPy/SciPy) computational pipeline that regenerates every number, table, and figure in this manuscript from source code. Under an illustrative, literature consistent parameterization (beta equals 0.30 per day, sigma equals 0.20 per day, gamma equals 0.07 per day, baseline CFR equals 0.50), we find that epsilon equals 0.30 reduces peak infectious prevalence by 28.7% (307.6 to 219.4 per 1000) and cumulative mortality by 19.1% (492.6 to 398.5 per 1000) relative to no cross protection, a difference that is statistically distinguishable from the null scenario under a two proportion z test (z equals negative 4.48, p equals 7.6 times ten to the negative sixth) and from an illustrative BCG heterologous protection proxy (z equals negative 2.57, p equals 0.010). A conjugate Beta Bayesian analysis of a hypothetical, not yet observed clinical efficacy contrast yields a posterior mean relative efficacy eta equals 0.656 (80% credible interval 0.417 to 0.862), and a variance based Sobol decomposition (Saltelli/Jansen pick and freeze estimators, n equals 1024 base samples) attributes 71.6% of peak prevalence variance to the transmission rate beta and only 5.5 to 8.4% to the hypothesized cross protection epsilon itself, a result we interpret, transparently, as evidence that the mechanism's epidemiological leverage is modest and secondary to outbreak containment fundamentals, not as confirmation of the underlying immunological claim, which remains entirely untested in humans. We specify quantitative, falsifiable immunological and clinical correlates (for example, less than 1.5 fold interferon gamma elevation in rabies primed peripheral blood mononuclear cells challenged ex vivo with filoviral glycoprotein pseudotypes) that would refute the hypothesis, and we lay out a phased preclinical and clinical roadmap. All code, data, and derivations are embedded in this self-contained document.