As Luigi Cattivelli tells it, the story of durum wheat starts more than 10,000 years ago, when some Neolithic farmers began cultivating a wild grass called emmer.

These farmers selected seeds that exhibited traits that made emmer easier to harvest and eat. Over generations, through selective cultivation and crossbreeding, human beings developed durum wheat and, later, bread wheat.

As small groups of people migrated from the Fertile Crescent, an area of the Middle East where the first farming is believed to have taken place, they brought seeds with them, further adapting the plants to the climates and conditions in new lands – including Italy, where durum became the source of pasta.

Cattivelli, an expert on the wheat genome, says the next chapter of the story deals with our immediate future; scientists must develop new varieties of wheat and other staple crops to meet the rapid pace dictated by climate change.

Cattivelli directs the Genomics Research Center in Fiorenzuola, which is part of the Italian government’s Council for Agricultural and Economic Research (CREA). Cattivelli and his colleagues, along with teams of crop geneticists from other parts of the world, are using high-performance computing in the Microsoft Azure cloud to try to unlock the genetic secrets of durum and other varieties of wheat. In the Pangenome Project, they are sifting through the genomes of about 40 varieties of wheat and its ancient ancestors for traits that would help the crop thrive in extreme conditions, be more efficient in use of natural resources and be resistant to disease and pests, reducing the need for fertilizers and pesticides.

It’s not just a question of pasta for Italians; it’s an urgent quest because growing enough staples like wheat, rice and corn is essential to human survival.

Wheat makes up about 20% of calories consumed by humans globally. And climate change is a direct threat to the production of crops globally, from drought and heat as well as torrential rains and other extreme weather events, such as the recent floods in eastern Spain.

Working together with Microsoft, CREA built a framework in the Azure cloud that eventually could house and analyze multiple petabytes of genetic data from the genomes of many varieties of wheat from multiple sources. (To get an idea of what that means, one petabyte could hold up to 2,000 years’ worth of digital music, if played continuously.)

Curtis Pozniak, a geneticist who directs the Crop Development Center at the University of Saskatchewan, Canada, is among the founders of the Pangenome Project.

“We’re generating petabytes of information that we need to filter down into something meaningful,” he says. “The only efficient way to do that is through cloud-based platforms where the same data can be shared with a whole range of experts at the same time.”

That data, which is stored in Microsoft’s Northern Italy Data Center Region, is then processed and analyzed in what is known as a “pipeline,” also housed in Azure. A pipeline is a series of data processing stages, in this case created with open-source coding. This particular genomic pipeline is designed to deal with billions of small sequences that have to be ordered to make the 14 chromosomes of the durum wheat genome. The pipeline is a tool that helps the scientists piece together that elaborate jigsaw puzzle.

This genomic puzzle can be seen and worked on by teams of scientists wherever they are in the world. Knowledge and information extracted from the genomic puzzle will be embedded in new varieties that will be made available to farmers in the coming years.

With the advantages of high-speed computing and effective collaboration, the speed of the research is greatly accelerated, says Pozniak, who is also a professor as well as being a wheat breeder.

He says he is excited for the researchers who are earlier in their careers because they have tools that weren’t available when he was working on his Ph.D. 20 years ago.

“It took me the better part of my Ph.D. to clone a single gene that was important for a wheat trait,” Pozniak recalls. “With the kind of data and analysis tools we have at our disposal now, we’re doing that in a matter of weeks or months. It’s an exciting time to be a scientist.”

Cattivelli has a personal history with wheat; he grew up on a farm in the Po River Valley about 20 kilometers, or 12 miles, from his office at the research center. His father grew wheat, among other crops.

The researchers at CREA are using a multidisciplinary approach called genomics to get a granular understanding of wheat (pun intended). Genomics combines biology, bioinformatics and information technology to analyze and interpret biological data. The tools are different, but the goals are the same as they have been for thousands of years – selecting traits to ensure the best possible harvest.

“Bioinformatics is only the latest chapter in the story,” Cattivelli says.

One of his colleagues at the Genomics Research Center, Primetta Faccioli, led the effort to create the system in the Azure cloud to store and analyze genomic data. She began her career as a “wet” biologist, working directly in the lab with plants. Now she describes herself as a “dry” biologist, working primarily with data.

Like Cattivelli, Faccioli grew up on a farm near the research center in Fiorenzuola and like him she fell in love with genetics – the story behind the plants her family was growing.

“Some years ago, we thought that the data production was more difficult than data analysis, but it’s not so,” she says. “We need both at the same level. The saying, ‘garbage in, garbage out’, is true. So, if garbage comes from the lab, bioinformatics produces garbage. So, we need to work together.”

The genome of wheat for bread was completed, to much fanfare, in 2017. But that is only one step. The genome, essentially a list of genes that makes up a life form, contains billions of DNA bases, and they are constructed in sequences (if you recall your introduction to biology class, those sequences are made up of an alphabet of four letters, A, C, G and T).

Those sequences are elaborate codes that spell out how a life form functions in the most intricate and minute detail. Wheat has a particularly elaborate genome – durum wheat has 10.5 billion bases and bread wheat has about 15 billion bases– that’s three times more than the human genome, the scientists at CREA say.

The idea is to find the genes that control specific traits in the plant, making it possible to create new varieties with the target traits and with greater speed, Faccioli says.

To return to the jigsaw puzzle metaphor, the Azure cloud creates the table where scientists can put all the pieces together. But with many people working at once, there must be quality control, reproducibility and portability, Faccioli says. (Before the cloud, many institutions were flying suitcases full of hard drives to each other to share data.)

Faccioli, Mario Giorgioni, an ICT specialist at CREA, and Wolfgang De Salvador, a Microsoft specialist in high-performance computing and artificial intelligence infrastructure, worked together to construct what is known as a pipeline.

They built this series of computing steps using a workflow orchestrator called Nextflow, created and supported by the Microsoft partner Seqera. This system in the Azure cloud made it possible for teams to work together using the same sequence of open-source programs to reach reproducible results. Giorgioni says they built this centralized research platform with tools that address the three main pillars of any high-performance computing system: fast shared storage, high-performance computing resources and fast network interconnects.

The infrastructure, built using Azure CycleCloud and leveraging Azure HPC services, allows researchers to easily scale computational resources as needed, he says. Azure Blob-based shared storage serves as the backbone, offering the necessary bandwidth and scalability for data-intensive research projects.

“Researchers have access to a wide range of computing resources to tackle the most intricate challenges in their daily work,” he says.

Nextflow-based pipelines are helping the scientists on the project convert raw data about the wheat genome into useful information, Faccioli says.

To put it in simple terms, Faccioli says, she often explains to visiting students that there is a huge difference between data and information. “I give them a number, and that is nothing but data. It doesn’t say anything,” she says. “But if I tell them that number is the phone number of this office, that is information.”

Elisabetta Mazzucotelli is one of the researchers in Fiorenzuola working on the Pangenome Project. Part of her work is combing through the genomes of many varieties of durum wheat and its ancestors to rediscover ancient but useful genetic traits.

As any plant is domesticated, there is a “bottleneck” effect where some genetic traits from the old plant are lost after crossbreeding and replicating a new variety.

“We need to find and record all the genetic diversity that is out there because now we are facing new challenges because of climate change,” she says. “There may be traits of disease resistance, or a capacity to thrive in the biome of the soil, that have been bred out of the current varieties of durum.”

Caterina Mare, a researcher at the Italian government’s Council for Agricultural and Economic Research (CREA) in Fiorenzuola, Italy, works to hybridize a rice plant. Photo by Chris Welsch for Microsoft.

Mazzucotelli says the supercomputing capacity in the Azure cloud allows her to manage and compare amounts of data that would have been impossible before. For her, it’s another tool in the pursuit of knowledge.

“I am very curious, and when can I see a plant that is green in a field that is full of other plants that are yellow and susceptible to disease, I am wondering why there is this difference,” she says. “And in most of the cases, the result is a difference at the genetic level. This means we can use this difference to bring innovation and to create a new solution. We can use the traits of this one green plant to make many, many more plants as green and healthy as this one.”