SOPHiA GENETICS is proud to announce the launch of SOPHiA DDM™ Dispatch. It represents an expansion of the SOPHiA GENTICS Integrated Solutions business area, which is a well-established model for working with partner labs to democratize data driven medicine by increasing access to next-generation sequencing and accompanying analysis. SOPHiA DDM™ Dispatch makes it even easier for partner labs to come online and operate independently.

With SOPHiA DDM™ Dispatch, our partner labs can handle orders directly with the clients, streamlining operations and improving efficiency.  Dispatch enables labs to take advantage of any unused capacity and creates an additional source of samples. For clients looking for sequencing services, SOPHiA DDM™ Dispatch offers increased data visibility, data ownership and simplifies the ordering process.

With SOPHiA DDM™ Dispatch and SOPHiA GENETICS Integrated Access, client institutions can get access to powerful insights without the setup costs. Clients send their samples to partner labs for sequencing, then the sequencing data is delivered to the institutions in their SOPHiA DDM™ account. This gives growing institutions better access to accurate analysis and allows them to retain ownership of their sequencing data. SOPHiA DDM™ Dispatch and SOPHiA GENETICS Integrated Access offer institutions more than just a report, by combining raw data ownership and ongoing access to powerful tertiary analysis.

Partner labs can easily begin offering service through either Integrated Solutions model, offering an opportunity to gain an additional revenue stream. Partner labs control the testing menu they offer and do not have to do data interpretation, since clients are enabled to do interpretation through SOPHiA DDM™. Furthermore, new tests can be seamlessly incorporated into your existing workflow.

Explore how the SOPHiA DDM™ Platform can help your institution today.

It’s not easy even for experts to pull useful insights from complex datasets. It’s even harder when the machine learning tools they’re using are only trained to work within a single nation, region, or more specifically, racial population.

Consider the recent popular Netflix documentary Coded Bias. MIT Media Lab researcher Joy Buolamwini discovered that facial recognition does not see dark-skinned faces accurately. She’s pushed for legislation in the US to govern against bias in algorithms. It’s a real problem that many in healthcare are now facing as the use of artificial intelligence in health data analysis becomes more and more prevalent and clinically useful. But the same push for diversity in machine learning reference for health data has been lacking throughout the industry. For this reason, and for the reasons you’ll learn from just one case study below, SOPHiA GENETICS has taken a decentralized and global approach for more than a decade.

A family without answers

A Moroccan couple living in Spain had suffered one of the worst human experiences, not just once, but twice. As they began to expand their new family, they had three pregnancies throughout the span of about one decade. Only one of their children, a healthy baby girl born in 2012, survived. One boy lived for 43 days, and one girl lived only for 20. The immense grief that this brings to parents is immeasurable, and the only thing nearly as horrible than the actual experience of loss is having a lack of answers – never knowing why.

After the death of their infant boy in 2019, the couple worked with regional medical experts in Spain to try and discover the root cause of what seemed to be neurodevelopmental disorders at birth. Researchers started by performing clinical exome sequencing and targeted gene analysis including a Microarray-based Comparative Genomic Hybridization test to look for any abnormalities that would possibly provide more clues. Much to their dismay, the results were inconclusive. The standard of care and testing used at the time were letting them down.

In 2022, they were able to perform more advanced testing after the death of their second daughter. This included TRIO analysis with the SOPHiA DDM platform. TRIO analysis takes clinical exome testing to a deeper level by analyzing data from the infants as well as both parents, giving researchers a better picture of familial, inherited genomic traits. The results finally confirmed that both parents carried a heterozygous frameshift variant. It was inherited by their infants as a homozygous pathogenic variant in the SMPD4 gene, causing neurodevelopmental disorder with microcephaly, arthrogryposis, and structural brain anomalies in the children. While identifying causative variants in recessive genes is challenging through conventional clinical testing, the SOPHiA DDM™ platform’s TRIO analysis enabled the researchers to overcome this limitation.

This family is now better informed about their risk of potential complications in future pregnancies. They’re able to work with genetic counselors to seek guidance regarding family planning that takes the anxiety out of the unknown.

Why diversity matters

In this case, it wasn’t just the type of testing that resulted in a successful discovery. Because SOPHiA GENETICS has worked on a global scale since inception (unlike more common approaches), the SOPHiA DDM platform’s machine learning algorithms have been better trained to analyze the health data of more diverse populations. Meaning, no matter where in the world this family had been trying to seek answers, their clinical researchers would have benefited from more accurate analysis of data that reflected who they are at the core of their genomes.

Every person is unique. We can clearly see this within the more than one million genomic profiles analyzed by our technology. This is why since day one SOPHiA GENETICS has adopted a global approach and placed it at the core of our company’s DNA. We strive to enable better research and health data analytics for all, no matter where they live or where they come from.

If you’d like to learn more about the SOPHiA DDM platform, contact us for a demo or learn more by exploring the many resources of our website.

Using a third-party sequencing service is an attractive option to institutions for whom the up-front capital investment of starting up a sequencing lab is an obstacle. There are many benefits to having access to a sequencing service, such as having predictable expenses and turn-around-times. However, in most cases there are significant challenges that remain in terms of data access and interpretation. Third party sequencing services often provide either a report for each sample, or just raw data. If a report is the only deliverable, institutions do not get access to the raw data and won’t be able to share it, archive it, or reanalyze it later. If raw data alone is delivered by the sequencing service, then institutions are on their own to find and test suitable secondary and tertiary analysis tools.

Gaining access to the raw data enables labs to perform data analysis in-house, for further analysis or to be aggregated with other samples for cohort-level investigation. Doing the analysis in house also gives the institution visibility into all variants present and detected within the sample, rather than just the reported variants. This visibility can lead to greater genomic discovery and increase knowledge of potentially relevant variants.

Doing data analysis in house can also help to support better decision making. Integrating the analysis platform into your institution’s existing health data management systems will allow decision makers to take a more comprehensive view of all the individual’s data which could help them consider more variants that might otherwise not be reported. Keeping interpretation in house will also allow institutions to consider the individual’s history. Breaking these data silos through integrations with native systems and bringing genomic analysis in house can support better decision making by focusing on the most relevant variants.

There are also cost benefits to doing the genomic analysis in house. Many sequencing services charge for the sequencing and the interpretation of the data. Doing the interpretation in house removed the associated costs, requiring expenses only for sequencing.

Decentralizing the analysis of genomic sequencing data can help institutions further adopt data driven medicine through ownership of their data. SOPHiA Integrated Solutions can help institutions bring genomic data analysis in house. With SOPHiA Integrated Solutions, samples are sent to one of our partner sequencing labs and the raw data is transferred back to your SOPHiA DDM account for analysis. This allows you to harness the analytical power of SOPHiA DDM to identify and report the relevant variants.

Precision medicine increases the efficacy of medical intervention by providing the right treatment to the right patient at the right time. Precision medicine not only results in better patient care, but also reduces some of the economic burden associated with challenging diagnostic odysseys and ineffective treatment plans. The advancement of healthcare technology has made the idea of precision medicine a reality.

One of the biggest leaps in precision medicine was the identification of actionable genomic variants, or biomarkers, for the targeted treatment of cancer. About a year ago, we celebrated the 20^th anniversary of the initial Human Genome Project publication, which allowed us to see the entire sequence of the human genome for the first time. Initiatives to close the gap in understanding the human genome continue today. For example, the All of Us project aims to collect over 1 million genomic sequences to increase the diversity of the current genomic knowledgebase. This research allows us to use machine learning and analytical algorithms to analyze genomic data and identify causative variants which can be used to select patients for targeted therapies.

Understanding our genomic blueprint made it possible to turn the idea of precision medicine into reality. Today we have access to precision immunotherapies and CAR-T cell treatments that can be used as alternatives to, or in combination with, traditional oncology care. Pembrolizumab is a prominent example of an immunotherapy for cancer patients with PD-L1 variations. Patients are tested for the specific PD-L1 biomarker to be considered a candidate for Pembrolizumab treatment. Another example of precision medicine in oncology care is CAR-T cell therapy, in which a patient’s own immune cells are modified with specific chimeric antigen receptors to help them destroy cancer cells in their own body. There are currently six FDA- approved CAR-T cell therapies for the treatment of lymphomas, leukemias, and multiple myeloma, and hundreds of ongoing clinical trials to target other types of cancers. You can learn more about CAR-T cell therapy from the American Cancer Society here.

However, cancer is an extraordinarily complex disease - to identify the most effective treatment, providers will have to move beyond a one gene to one drug mentality and rely more on comprehensive multimodal patient data.  Patient genomic profiles can vary greatly -, a recent study found over 5000 unique mutations in 628 cancer-associated genes across 54 tumor types¹. Multimodal patient data aims to go beyond genomics, to consider more comprehensive biomarkers from imaging, proteomics, transcriptomics, epigenomics, and phenotypic and medical health information. By combining these -omic profiles we get a holistic view of the individual and we can create more fractionalized and specific cohorts of individuals with better-defined disease states. These highly specific disease states can be used by pharmaceutical companies to develop new, more targeted, therapies, increasing efficacy and resulting in better patient outcomes.

Similarly to how genomic information became actionable, the use of other -omic data will also require extensive data collection and knowledgebase creation to identify relevant variants for specific disease states. Through machine learning, we will be able to extract trends from relevant cohorts to provide data beyond what is available in current reference databases, to generate knowledgebases with more global inclusion, diversity and knowledge. Providers will be able to leverage this data to classify patient profiles and to confidently identify treatment plans using predictive outcome capabilities created from similar patient profiles. Advances in cloud-data storage and machine-learning algorithms make this future more of a possibility. The field of precision medicine continues to evolve but remains steadfast in its goal to provide the right treatment to the right patients at the right time. Increases in the analysis and utilization of multimodal patient health data will help healthcare institutions achieve that goal.

References

1. Zhao Y., et al. (2021) ‘PO2RDF:Representation of Real-world Data for Precision Oncology Using Resource Description Framework’, BMC Medical Genomics, in review.

As the demand for genetic testing continues to grow, more institutions are looking for ways to incorporate next generation sequencing (NGS) to augment their in-house capabilities. Increased access to genomic testing can help fulfill the promise of making precision medicine accessible to all, but the upfront investment required for institutions to start their own sequencing labs remains a barrier. To overcome this challenge, institutions can send their samples to service laboratories for testing and analysis, and this has traditionally been known as using a send-out service.

When using a traditional send-out service, samples are sent to a reference lab and the institution receives a report with the relevant findings. Outsourcing to reference labs does increase access to NGS tests, however, it can come with its own challenges. With traditional send-out services, institutions can struggle with turnaround time and a lack of data ownership or customizable reports. Turnaround time is crucial when trying to make decisions in a short timeframe. Access to the original sequencing data allows institutions to not only increase their knowledge of genomic data interpretation, but also enables the ability to aggregate data for further analysis that may be relevant for their region or field. The reporting needs of each institution can vary greatly, and not all reference labs can provide the flexibility in reporting needed to meet these needs.

SOPHiA Integrated Solutions presents an alternative to using a traditional send-out sequencing service. With SOPHiA Integrated Solutions, users send their specimens or nucleic acids to one of the sequencing partners within our global network. The chosen partner performs the wet-lab and sequencing work, and then the original sequencing data is securely transferred to the user via SOPHiA DDM™. This allows the institution to do their own interpretation with the help of powerful algorithms within the SOPHiA DDM™ platform. Through SOPHiA Integrated Solutions, institutions maintain ownership of the data which can be used for future aggregated analysis. The SOPHiA DDM™ platform also provides access to reporting tools with unlimited customizable reports so information can be communicated clearly and efficiently.

Instead, it can accelerate and validate researchers’ discoveries for faster application of their hard work to improve patient care. In healthcare, we see a promising step toward better patient outcomes thanks to advancements in machine learning.

AI and NGS

There’s a reason why it took so long for the first human genome to be sequenced. It’s not an easy thing to accomplish without a “roadmap.” When researchers analyze genomic samples, they’re often overwhelmed with countless panels to run, limited time and resources, and a whole ton of data that gets pulled from every single sample. With artificial intelligence, that first “roadmap” discovered in the Human Genome Project isn’t such a long journey anymore. Algorithms can help quickly identify biomarkers within a person’s genomic data that can assist clinicians shed light on many medical mysteries.

Artificial intelligence has been adopted more widely for clinical trials and patient care in recent years thanks to the evolution of NGS (Next Generation Sequencing). Even when it’s not being utilized as the key player to gather important new insights, machine learning applied to Next-Generation Sequencing can better organize and identify variants of interest hidden among background noise. Algorithms trained on specific data points discover and flag biomarkers in new, more efficient ways, down to the exon level. More in-depth output supports better informed decisions.

AI in radiology

Imaging is one of the first tests performed when a patient begins the diagnostic journey. Traditionally, an expert is trained and educated to specifically search for and identify concerning areas or segments of a given medical image. Artificial intelligence in radiology goes well beyond what the human eye can see by analyzing the available data, not just the image.

Through 3D segmentation and visualization, AI can pinpoint concerning features in medical scans faster and more efficiently than the human eye. Recent studies even suggest that AI for radiomics could be used in preliminary interpretations of chest radiographs to address the scarcity of resources, improve accuracy, and reduce the cost of care.

Democratizing AI-powered data for all

Trustworthy AI requires data volume and diversity. The more unique data that the algorithms can train upon, the more accurate they become in searching for relevant biomarkers within a sample. This is why SOPHiA GENETICS created a universal platform that can adapt and evolve with its users. A global community of more than 780 health care institutions have already supplied the SOPHiA DDM™ platform with relevant data, analyzing more than 770,000 genomic profiles. Standardized, pseudonymized, and aggregated data improve our machine learning algorithms to deliver variant detection, analysis, and interpretation that empowers researchers beyond manual investigation.  Learn more about SOPHiA DDM^TM, combining data from AI in Genomics and Radiomics with our technology by clicking here.