The science of COVID vaccine development certainly doesn’t lack for technical jargon. But the underlying mechanisms of this grand medical endeavor could very well spell the difference between success and failure.

There are more than 65 coronavirus vaccine candidates in various stages of development around the world, including some 50 in human studies. The four most prominent ones getting attention in the United States, with large-scale phase three clinical trials, stem from Moderna, Pfizer, Johnson & Johnson, and the U.K.’s AstraZeneca.

And these companies are attempting to pull off the near impossible: Take a vaccine development process which can take up to a decade and compress it into a year-and-a-half timeline. And infectious disease vaccines don’t exactly have the greatest of track records (just 34% succeed, according to one study by MIT).

These vaccine candidates have wildly divergent action mechanisms. To put it more simply: They affect your cells and immune system in very different ways. Here’s how some of that science works.

Moderna

Two of the buzziest COVID-19 experimental vaccines use something called messenger RNA (or mRNA) technology.

Moderna is a bit of an odd story in this field. Its COVID candidate, currently dubbed mRNA-1273, could well become the first mRNA vaccine approved ever (and the company’s first product to make it to market.)

Ok so… What the heck is mRNA? Vaccines are what are called biologics. They depend on actual biological material because they need to induce an immune response which builds antibodies and then, hopefully, builds up protection against a pathogen.

With mRNA, you use this kind of biological messenger to instruct cells to carry out that critical task of becoming a natural drug maker within your body. That is, you turn your very biology into a personalized manufacturing site.

“RNA vaccines work by introducing an mRNA sequence (the molecule which tells cells what to build) which is coded for a disease specific antigen, once produced within the body, the antigen is recognized by the immune system, preparing it to fight the real thing,” according to the University of Cambridge.

There’s another twist to that science: It requires ultra-cold refrigeration in order to preserve its various components. And ultra-cold means ultra-cold, to the tune of negative 70 or 80 degrees Celsius, or about negative 112 degrees Fahrenheit. This may require liquid nitrogen or specialized carrying equipment, and Moderna’s specific candidate requires a two-dose regimen.

Pfizer/BioNTech

Pfizer’s story is similar to Moderna’s, although the company has a significantly bigger footprint in the pharmaceutical world as a $51 billion-plus revenue firm (and the logistics and manufacturing advantages which come with that, including a specialized case to ship its own vaccines.)

But, just as with Moderna, Pfizer will face the issue of storage and delivery given the mRNA technology its experimental therapy relies on. It literally created a device which can track the temperature and exact location of any dose being shipped across the country or the world.

“The logistics of medicines and distribution of medicines are always very complicated, because there are always storage conditions,” said Pfizer CEO Albert Bourla during a virtual Fortune conference earlier this month. “And we knew that we had to move very fast. So we wanted to make sure that we can distribute by plane, we can distribute by any type of vehicle without needing refrigeration individually.”

Much like the Moderna tech, Pfizer and its partners’ vaccine would seek to turn your cells into antibody-producing machines, and initial results from the company suggest it’s effective at producing so-called neutralizing antibodies that can, at least temporarily, shut the virus down in its tracks. We should know a whole lot more by late November.

AstraZeneca

AstraZeneca, the British pharmaceutical giant working in tandem with the University of Oxford, is working on a therapy called AZD1222. The therapy is closer in line with what Johnson & Johnson’s pharmaceuticals arm, Janssen, is working on—a vaccine which latches onto the so-called “spike” protein of the coronavirus.

This is what’s called a “non-replicating ChAdOx1 vector vaccine.” It attacks the S-protein, which allows for the entry of a virus into a cell.

That’s a pretty critical line of defense against a pathogen given that this is how virus replicate, sticking onto these cells and turning them into hosts (this is one of the essential distinctions between viruses and bacteria, the latter of which can replicate on their own.)

So, rather than turn cells into antibody-producers, this approach prevents the actual entry of the virus.

Johnson & Johnson

Johnson & Johnson has expressed some optimism about its coronavirus vaccine candidate, especially because it claims it can be effective with just a single dose rather than two administered within a number of weeks.

A “single immunization with Ad26.COV2.S showed no detectable virus in the lower respiratory tract after exposure to SARS-CoV-2,” the firm said earlier this year.

Like the AstraZeneca candidate, J&J targets the coronavirus spike protein via an inactive common cold vector that makes its way into cells (certain versions of the common cold are among the same family of pathogens as the coronavirus and other respiratory diseases).

On the logistics side, this particular vaccine may not need the ultra-cooling that Moderna and Pfizer need, possibly being able to be stored in more conventional refrigerators for a longer time.

International candidates

This list is an America-centric one; there are plenty of other vaccines in various stages of development, according to outlets such as RAPS and the New York Times, with potentially hundreds internationally in preclinical and clinical stages.

From CureVac to Japan’s Agnes and Takara Bio to Arcturus, these companies are relying on everything from synthetically made antibodies, to spike protein-blocking compounds, to messenger RNA to tackle the bug.