Juan glanced at the screen, which showed a complex map of the rat genome. He smiled and knew exactly why the intern was perplexed. For a well-documented genome such as Rattus norvegicus, the common brown rat, the screen should have been awash with green squares, indicating identified genes. Instead, the screen was filled with large splotches of red—gene sequences that didn’t match anything in the database.

“Are you sure the genome we’re working from isn’t corrupted?” Mike asked. “I’ve done this before at school and I’ve never seen so much unidentified data.”

Juan couldn’t tell the kid the truth. None of the interns knew that the genome for specimen 153 represented the results of five thousand years of simulated evolution. But he was prepared; he’d told this lie before.

“Mike, I’m sure you’re familiar with hidden Markov models. The software we’re using leverages HMM to identify the location of genes within the genome I supplied to you. It’s not uncommon for mutations to occur in these samples, for reasons I can’t disclose. And even the most extensive gene database in existence would mark these mutations as unidentified. The software is only as good as its database.”

Juan tapped on one of the red boxes on the screen. It zoomed in to show the signature of what the analysis had found. “Still, by flagging the changed sequences, the software had done part of the work. We can tell that this sequence is functional, but not what its function actually is. Our job is to fill in the blanks. We can only do that with in-vivo experimentation.”

The intern glanced at the far wall, which had a large window looking into the neighboring lab. The one filled with the rat’s enclosures. He frowned. “So, do you suggest doing a knockout assay—”

“No.” Juan tapped again on the computer screen, which reverted to the map of green and red boxes. “To understand what these functional sequences do, we’ll need to modify our test subjects. The protocol is simple. You’ll isolate the identified sequences and insert within them the synthesized sections that the program noted as unidentified. We’ll then take those constructs and insert them into the hosts with a modified viral agent.

“Believe me, I know it seems like a lot of work. I’ve been winnowing out the meaningless differences in the genome for a few years. But one or more of these sequences in combination may yield the key to making humanity impervious to some types of cancer. All the puzzle pieces are in front of us. We just need to figure out how they all fit.”

Mike’s eyes widened, and he gulped. “So, we’re going to need to inject the rats with the modified viral agents.” He looked queasy.