A Mathematical Analysis and Prediction of Ebola Virus: Past, Present and Future — Comparison with the Research from Local Health Authorities, WHO and Harvard University

Since its first outbreak in 1976, cross-regional mutation and transmission of Ebola virus have rendered epidemic forecasting a challenging task in global public health and viral evolutionary dynamics. Major international mainstream analytical frameworks include the SEIRF compartmental statistical model developed by the World Health Organization, the genomic stochastic transmission dynamics model from Hassan Bis’s team at Harvard University, time-series fitting models adopted by local health authorities in endemic regions, and the viral mutation statistical system maintained by NCBI GenBank and Nextstrain. Conventional mainstream paradigms are characterized by post-hoc data fitting and empirical parameter assignment. Their research scope is largely confined to retrospective analysis of completed outbreaks and backtracking of observed gene mutations. Lacking fixed underlying mathematical constraints, these frameworks cannot support prospective forecasting or quantitative extrapolation of long-term mutation trends, and suffer from inherent limitations in prediction timeliness, mechanistic completeness and quantitative stability. This work presents an original multi-factor coupled mathematical system under global steady-state constraints. A fixed global steady-state reference constant is defined. Four dimensionless core factors are introduced: macromolecular component ratio, structural degree, macromolecular torsional index, and systemic balance index. A two-layer bias correction algorithm is constructed to establish a mathematical mapping between microscale steady-state distortion of viral macromolecules and macroscale epidemic size. The manuscript consists of three parts: retrospective mathematical analysis, verification against observational data, and trend extrapolation. All input data for calculations, benchmark metrics and mutation statistics are obtained from publicly available authoritative sources: WHO, the Harvard research team, the Ministry of Health of the Democratic Republic of the Congo, NCBI GenBank and Nextstrain. The model rules were pre-defined; no post-hoc parameter tuning or fitting was performed. Retrospective calculations were carried out for all terminated Ebola outbreaks from 1976 to 2025. The model was further validated against observational data of the 2026 Bundibugyo Ebolavirus outbreak in the Democratic Republic of the Congo, followed by extrapolation of viral steady-state trends. The model yields a relative prediction error of only 1.04% for this outbreak, outperforming mainstream frameworks in quantitative accuracy, prospective capacity and mechanistic integrity. Model calculations indicate that Ebola viral variation is not purely random mutation as commonly recognized in academia. Instead, it constitutes a continuous, ordered evolutionary process constrained by macromolecular steady-state mechanisms, with quantifiable, reproducible boundaries for mutation and epidemic scale, enabling prospective estimation. This study breaks the conventional paradigm of “post-hoc statistics and retrospective review” for severe infectious disease research, establishing an integrated prospective quantitative framework for retrospective analysis, real-time verification and trend forecasting. The full set of equations has a fixed structure and reproducible calculation workflow. It provides a novel mathematical paradigm and quantitative reference for investigating evolutionary mechanisms of Ebola and other highly variable RNA viruses, quantitative epidemic assessment and prospective risk warning.