In the high-stakes race against Ebola, scientists are fighting back with an unexpected weapon: the structural similarity between a COVID-19 pill and a flu drug.
The Ebola virus, with its terrifying mortality rate of 25% to 90%, has long represented one of humanity's most formidable viral foes4 . For decades, management of this brutal disease relied entirely on supportive care—fighting symptoms without directly attacking the virus7 . The 2014-2016 West Africa outbreak, which claimed over 28,000 lives, starkly exposed the critical need for effective treatments4 . Today, in a remarkable turnaround, scientists are using advanced computational methods to discover promising new treatment approaches by finding hidden connections between existing drugs. This article explores how cutting-edge atom-pair fingerprinting and scaffold network graph algorithms are revealing unexpected similarities between medications, potentially opening new frontiers in the battle against Ebola.
Before understanding the treatment breakthroughs, it helps to know what we're fighting. Ebola virus disease (EVD) begins with deceptively ordinary symptoms—fever, headache, muscle pain, and sore throat5 . But it can rapidly progress to severe vomiting, diarrhea, impaired organ function, and in some cases, internal and external bleeding5 . The virus achieves this devastation through a mere seven structural proteins that perform astonishingly complex functions.
Interactive visualization of Ebola virus structure with glycoprotein spikes
Until recently, doctors could only support patients through the brutal course of the disease—managing hydration, electrolyte balances, and organ function while hoping the patient's immune system could mount a defense4 . The landscape changed dramatically in 2020 when the FDA approved the first specific Ebola treatments: Inmazeb (a cocktail of three monoclonal antibodies) and Ebanga (a single monoclonal antibody)1 4 . These groundbreaking therapies marked a turning point, demonstrating that direct antiviral strategies could succeed against Ebola4 .
Developing new drugs from scratch typically takes a decade and costs $2-3 billion3 . This daunting timeline becomes ethically unacceptable during deadly outbreaks. Consequently, scientists have turned to drug repurposing—finding new uses for existing medicines—as a strategic alternative.
"The primary goal of drug design is the discovery of new compounds with desirable pharmacological properties," note researchers in Computational Biology and Chemistry. "Chemical scaffolds have a long history in chemistry as a common core structure that characterizes a group of individual molecules as substructures"3 .
The foundation of this approach rests on a simple but powerful principle: structurally similar molecules often share similar biological properties3 . This concept drives the search for drugs that might work against multiple diseases. Two computational methods have become particularly valuable in this quest:
These methods break down molecules into their core structural frameworks, creating family trees of related compounds that help researchers understand which parts of a molecule are essential for its activity3 .
These approaches allow scientists to navigate the vast universe of chemical structures efficiently, finding unexpected connections between drugs that might appear unrelated at first glance.
In a 2022 study published in Computational Biology and Chemistry, researchers applied these computational methods to drugs with potential activity against Ebola2 . Their investigation yielded a surprising discovery: significant molecular structure similarity between favipiravir and molnupiravir2 —two antiviral medications already familiar for other viral threats.
The research team followed a systematic computational approach:
They gathered structural information for FDA-approved and experimental anti-Ebola drugs from public databases including PubChem and DrugBank2 3 .
Using atom-pair fingerprint technology, they quantitatively compared the two-dimensional structures of these medications, calculating similarity scores based on their molecular fingerprints2 3 .
Through scaffold network graph methods, they broke down promising drug molecules into their core structural components, following the definition by Bemis and Murcko that describes "a scaffold as a combination of rings and atoms that connect them"3 .
They created graphical representations of the molecular architecture and core structures of the most promising candidates2 .
The computational analysis revealed that favipiravir and molnupiravir share unexpected structural similarities that might contribute to antiviral activity2 . Both drugs belong to a class known as nucleoside analogs, which work by mimicking the building blocks of viral genetic material2 3 . When the virus attempts to replicate, it mistakenly incorporates these imposters into its RNA, leading to fatal errors in the genetic code that ultimately stop viral replication2 .
| Drug Name | Primary Approved Use | Similarity Score |
|---|---|---|
| Favipiravir | Influenza |
|
| Molnupiravir | COVID-19 |
|
| Remdesivir | Ebola, COVID-19 |
|
| Scaffold Type | Representative Drugs | Mechanism |
|---|---|---|
| Nucleoside analog | Molnupiravir, Favipiravir | RNA replication errors |
| Monoclonal antibody | Inmazeb, Ebanga | Blocks virus entry |
| Synthetic siRNA | TKM-Ebola | Gene silencing |
| Factor | De Novo Drug Development | Drug Repurposing |
|---|---|---|
| Timeline | 10+ years | 1-2 years |
| Development Risk | High | Substantially reduced |
| Safety Profile | Unknown | Already established |
| Cost | $2-3 billion | Significantly lower |
The scaffold network analysis further illuminated why these drugs might work against diverse viruses. Both medications contain core structures that target a fundamental aspect of viral biology: the replication machinery2 3 . This explains why a drug developed for influenza (favipiravir) and another for COVID-19 (molnupiravir) might find application against Ebola—they all attack RNA viruses that depend on similar replication processes.
Behind these computational discoveries lies a sophisticated array of research tools that make such analyses possible:
Tools that deconstruct molecules into core frameworks, creating visual representations of structural relationships between compounds3 .
Programs that compute Tanimoto coefficients, Dice indices, and other metrics to quantify structural similarities between molecules3 .
An open-source cheminformatics toolkit that implements various topological fingerprint methods for molecular analysis8 .
While the structural similarities between favipiravir and molnupiravir are scientifically compelling, the researchers caution that their findings represent a starting point rather than a finished solution2 . The combination they identified requires "further research for treating Ebola," with laboratory validation and clinical trials needed to confirm whether these computational predictions translate to real-world efficacy2 .
Current approved therapies like Inmazeb and Ebanga specifically target the Zaire species of ebolavirus1 4 , but the 2022 outbreak in Uganda was caused by the Sudan species—against which these treatments may be less effective4 . This underscores the urgent need for broad-spectrum approaches that could work across multiple Ebola species.
The promising combination of favipiravir and molnupiravir represents just one application of these powerful computational methods. As the authors note, understanding "the core structure(s) of medication molecules effective against the Ebola virus, their inhibitors, and the chemical structure similarities of existing pharmaceuticals" provides a foundation for designing future drugs and drug combinations2 .
The story of atom-pair fingerprints and scaffold network graphs in Ebola research represents more than a technical achievement—it signals a fundamental shift in how we approach deadly infectious diseases. By using computational methods to find hidden connections between existing drugs, scientists have potentially compressed years of traditional drug development into months of intelligent analysis.
As the researchers conclude, "The combination of molnupiravir, the first licensed oral medication candidate for COVID-19, and favipiravir, employed in other viral outbreaks, should be further researched for treating Ebola"2 . This statement captures the new paradigm in infectious disease response: leveraging advanced computational tools to find unexpected connections between existing medications, potentially yielding rapid treatment strategies for the next outbreak before it becomes a pandemic.
In the enduring battle between human ingenuity and viral evolution, these molecular blueprints may ultimately provide the key to turning the tide against Ebola and other emerging threats. The work continues, but the computational tools now available offer unprecedented hope for a future where no outbreak finds us defenseless.