About one in 100 people is affected by such characters. For some, the impact may be slightly moderate; in others, it’s life changing. Take for instance the story of Mira and Sai, a young couple who lost two sons, both under 18 months, to a mysterious disease the physicians and geneticists couldn’t diagnose.
The boys, Som and Rom, had prominent eyes, small chins and full cheeks. Could there be a clue in those features? Once the genomes of Som and the parents were sequenced, a list of variant characters was ready for investigation. Whittling them down to the organs that were affected—facial structures, lungs (respiratory problems) and so on, the investigators ended up with a few hundred genes, which might hold the secret to the mystery. Narrowing the gaze further, through some thought experiments that took them back 500 million years, they arrived at an errant character. But it was uncharted territory—nowhere in genomics literature had that genetic anomaly been reported. And while Hariharan’s team pored over tonnes of data and scoured the worldwide web of scientific literature, time was running out. The couple was expecting their third child and was desperate to know if the baby would meet the same tragic fate as the two boys.
“In some cases, the defective characters are known to the doctors but in this case, it was an unexpected gene. You had to convince the doctor who herself was not sure and was in a dilemma if she should use our information to take a decision to terminate the pregnancy or not,” says Hariharan, relieved that the doctor relied on their results. “When it comes to patients, you can think of us as a glorified reporting service, which says—here are the things, which indicate we have found the answer; here are the things, which indicate we are in the realm of the unknown. And the doctor took a call at that point which was to do nothing. And it turned out fine.”
Moves and countermoves
Genomic Quirks: The Search for Spelling Errors is a collection of stories that shed light on an entire variety of diseases whose underlying genetic causes are not easily identifiable. Each story is a different medical phenomenon—eye, heart, blood and so on. Each required a different algorithm after the team realised that something was wrong. And each story has a different genomic phenomenon. If one is a cut and paste case, of a few characters getting cut in one part of the genome and being pasted in another, another is where a whole chunk of the DNA has been removed.
It’s also a story of a genomics war room: Amidst hundreds of cases coming in and all hell breaking loose, how do you find the right answer? The preface captures it to an extent but each story peels off the complexity a little more. The initial understanding is limited even for the best technologists. But like a detective piecing together clues and evidence, Hariharan’s team builds a plan to nail it. In every case.
“Our hunt for the culprit is now nearing its end. We look in our shortlist of rare missense and nonsense variants for those, which appear in both versions of Dia’s genome. And how many candidates do we find fitting this bill? Unfortunately, none!” writes Hariharan in the second story. Clearly, he is inspired by Sherlock Holmes. The similes, the imagery and the plot movement are all Holmesian. The maestro is quoted as well: “Once the impossible has been eliminated, whatever remains, however improbable, must be the truth.”
The context in the book for Holmes’ words is this: A genetic variant has been found but is it really the culprit? “There’s a catch in Holmes’ claim. Unless we have considered every possibility, we cannot eliminate all that is impossible. Have we really considered every possibility?” writes Hariharan. Step by step, he takes the reader through the elimination process. It’s logical and easy to understand even as the language is fraught with the inescapable genetics jargon.
“No problem is too hard to go after, for Ramesh. He finds a systematic way of breaking them down, be it in business or technology,” says Vijay Chandru, co-founder and chairman of Strand. When a professor of computer science at the Indian Institute of Science (IISc) in Bengaluru, Chandru had heard about Hariharan’s work at the Courant Institute of Mathematical Sciences in New York. It was the mid-1990s, and the connection between algorithm and biology had already been established. The Human Genome Project was in progress. “Ramesh had one of the best algorithms for matching strings of [abstract] letters, which could be used in genetics [where the DNA is a string of letters A, T, C and G],” says Chandru.