支付宝小程序或于下周正式发布 将与微信刚正面！

Data analytics technology is a powerful enabler for genome sequencing.

With only one case in history, Ribose 5-Phosphate Isomerase (RPI) deficiency is the rarest genetic disease in the world. It affects nerve cells, causing weakness, numbness, and pain in the hands and feet, as well as disrupting important bodily functions such as digestion, urination, and circulation.

While identifying rare diseases such as RPI deficiency is far from straightforward, genome sequencing can help. Indeed, more broadly speaking, the approach is capable of uncovering millions of mutations, allowing doctors to better understand and treat a wide range of diseases.

That said, healthcare facilities need to be able to handle large amounts of data to make genome sequencing possible. And Huawei's High-Performance Data Analytics (HPDA) Solution is helping to do just that.

Uses of Genome Sequencing

Genome sequencing involves breaking Deoxyribonucleic Acid (DNA) into multiple pieces and arranging the fragments in sequences. Medical experts then use this data to understand the genetic composition of viruses, for example.

Just such a method was key to supporting public health efforts during the pandemic: it was the understanding of the genetic code of COVID-19 that helped scientists to rapidly design effective vaccines.

Researchers are also able to compare individual gene sequences to identify anomalies in a person's genetic code. Variants in DNA may pose health concerns and have been linked to several types of cancer and developmental disorders, the influential publication American Scientist noted.

Huawei is actively working with West China Hospital (WCH), headquartered in the Chinese city of Chengdu and one of the largest single-site hospitals in the world, to study how behavioral and environmental factors affect genes. This approach, known as multiomics, gives doctors a better understanding of how molecular changes contribute to development and disease.

They can then use this information to customize medical treatment. "Innovation will accelerate the widespread application of big data in precision medicine," Dr. Yu Haopeng, a data scientist at WCH, said.

In 2020, for example, Huawei partnered with the hospital to analyze the genes of 100,000 Chinese patients with rare diseases. The project allowed the hospital to create personalized medicine, pinpoint gene mutations, and predict a person's susceptibility to disease and drug response.

Challenges in Genome Sequencing

Huawei’s HPDA Solution has helped WCH shorten the hospital's genome sequencing analysis time from 24 hours to just 7 minutes. Achieving such a staggering reduction was no easy feat. Indeed, success hinged on overcoming three main challenges.

To begin with, when data from each step of the genome sequencing process is stored in different places, scientists risk losing data during the transfer process. Any network interruption, for instance, could spell disaster, negatively impacting the efficiency of analysis. A solution that allows users to access the same piece of data over the different stages of genome sequencing is an ideal solution, here.

Next, genome sequencing takes up a lot of storage space. Huawei estimates that a person's genome data is around 100 GB in size. To put that into context, it's the equivalent of approximately 208 days of 24/7 Internet browsing, according to British Internet Service Provider (ISP) Broadband Wherever. Sequencing efforts, nonetheless, need to cater to the hundreds of thousands of patients a hospital like WCH serves.

Finally, analyzing all this data requires intensive computing resources. The process involves massive data transmission, placing huge pressure on network bandwidth.

Three Problems: One Solution

To overcome these three clear challenges, genome sequencing requires a platform that can seamlessly integrate the data from each step, support high network bandwidth and computing power, and store large amounts of data. HPDA appears to be the answer to all three.

Huawei's powerful computing system allows users to analyze data across the different processes of genome sequencing without the need for migration. This is thanks to essential technologies like lock-free data structures. Simply put, if one process is interrupted in the middle of genome sequencing, other processes are not locked out from doing their jobs and can proceed as usual. There is no need to put data on hold while waiting for all processes to reboot.

Data migration is also no longer necessary with multi-granularity disk space management. This is an algorithm that pre-allocates disk space to improve the performance of storage devices. Instead of figuring out where data should reside and how it should travel from one process to another, technology has it all figured out.

Stable high bandwidth and higher storage capacity have also made quicker genome analysis and testing possible.

Huawei says that a storage system should deliver at least 6 GB/s of data to ensure its accuracy, completeness, and consistency. Huawei's HPDA Solution offers 32 GB/s bandwidth, more than five times the minimal level designated.

And while it used to take 13 years to fully sequence the genome of a single person, with the help of HPDA, labs can now achieve this feat within a single day.

But supporting such a large database requires high power output, and this is no small fiscal undertaking. However, the 5-year cost of owning Huawei's solution is 61% lower than that of an average computer data storage server, analysts Enterprise Strateg Group (ESG) have said.

Developments in the genome sequencing industry pose new challenges to high-end storage, but HPDA can overcome them. And a storage solution can help hospitals to develop treatments faster, supporting better patient recovery — which is the ultimate aim of healthcare.

Disclaimer: The views and opinions expressed in this article are those of the author and do not necessarily reflect the official policy, position, products, and technologies of Huawei Technologies Co., Ltd. If you need to learn more about the products and technologies of Huawei Technologies Co., Ltd., please visit our website at e.huawei.com or contact us.