SAN FRANCISCO (GenomeWeb) – A group of Stanford University researchers has spun out a startup that will be headquartered in California, but also have Hangzhou, China-based operations, and that is focused on using exome sequencing to diagnose neurological disorders, with autism a particular focus.
Mike Snyder, a cofounder of SensOmics and director of Stanford's Center for Genomics and Personalized Medicine, said that the company has licensed technology developed at Stanford and has partnered with the Children's Hospital at Zhejiang University School of Medicine in Hangzhou for its initial research.
The company, which currently has around 10 full-time employees, has received seed funding from Asset Management Ventures and iSeed. It has also received a grant from the local government in Hangzhou that it will use to build its sequencing and computational facilities and expand its team. Part of that grant money has also been used in the collaboration with the Children's Hospital at Zhejiang University.
SensOmics also plans to raise additional funds in a Series A round later this year.
The company's ultimate goal is to enable earlier diagnosis of autism so that interventions can be started earlier.
"We see this as an area of very unmet need," Snyder said. In addition, "it's well established that earlier diagnosis offers a better outcome," Zhaolei Zhang, a cofounder and bioinformatician at the University of Toronto, said. Traditional methods are not able diagnose the disorder until an average age of four years, he said. "We're hoping to lower the average age of diagnosis by one to two years."
To do this, the firm plans to build on research conducted at Stanford that identified molecular networks and neuronal protein complexes linked to autism. It has filed for a patent on its systems biology method through the Stanford Office of Technology Licensing and plans to file additional patents on the machine learning algorithms it is developing. It is also building up a proprietary database of autism-related genes and variants, according to Snyder.
According to Jingjing Li, a cofounder of SensOmics and an instructor in the department of pediatrics at Stanford, the firm is working with medical practitioners to build machine-learning models that will integrate genomic information with other clinical data to improve the diagnostic yield of exome sequencing and to "find clues for undiagnosed diseases."
In a pilot study with researchers from the Children's Hospital at Zhejiang University, the SensOmics team used exome sequencing to analyze 10 patients with diseases that were unable to be diagnosed using standard clinical examinations. They identified the molecular etiology of four of those patients, including one patient who had a hybrid manifestation of two diseases: a developmental disorder known as Smith-Magenis syndrome — which typically results in intellectual disability, distinctive facial features, and sleep and behavioral problems — and neurofibromatosis.
SensOmics now plans to extend its collaboration with the hospital, focusing specifically on patients with suspected autism. For this collaboration, the firm has developed a machine-learning tool called AutProfiler, which identifies autism spectrum disorder-related genes.
Qiang Shu, president of the children's hospital, said, "Timely diagnosis of children with neurological diseases is a major challenge facing our doctors." As such, one goal of the collaboration is to "introduce [genomic] technology to the hospital" in order to improve diagnosis rates, including diagnosing children earlier.
Shu said that if the initial work with SensOmics proved successful, he could envision a future in which SensOmics' sequencing and disease-risk prediction pipeline is used before children start displaying clinical symptoms. For instance, he said, in cases where an initial newborn screening test has abnormal results or even in children who are at high risk of autism due to parental age, he said. The hospital has a large newborn screening center where it screens around 500,000 babies each year using mass spectrometry, Shu said. The hope is that by implementing exome sequencing as a follow-on test when NBS is abnormal and genetic disease is suspected, it could "help us discover the etiology of many rare diseases," Shu said.
SensOmics is also working on forging collaborations with other hospitals in China to implement its sequencing and machine-learning technologies and also plans to expand to other neurodevelopmental diseases. Aside from AutProfiler, SensOmics is developing a tool called OmniProfiler, which will predict risk for more than 200 diseases, including inherited metabolic disorders.
SensOmics' business model will rely on both research grants and partnering with hospitals. For instance, Zhejiang University funds some of the research work with SensOmics. In addition, the company plans to develop its tools into products, Li said, the first of which it would announce by the end of the year.
Currently, SensOmics' work does not need approval by the China Food and Drug Administration, Li said. Although, in the future, the firm may launch products for which it will require such approval, he added.
SensOmics is not the only group to focus on using genomics to enable earlier diagnosis of autism. A Canadian project known as Individualized Treatments for Autism Recovery using Genetic-Environment Targets (iTARGET) is analyzing the genomes, microbiomes, metabolomes, and phenotypes of more than 1,500 autism patients and their families to identify links between genetics and environmental factors that contribute to autism with the goal of developing diagnostic tools that would enable earlier diagnosis as well as biomarkers that are suggestive of treatment.
The iTARGET consortium is working with another autism-focused project, MSSNG, a long-running collaboration between Autism Speaks, the Hospital for Sick Children in Toronto, and Google, that is developing genomic and behavioral data resources about autism.
Kaiser Permanente has also started a project to enroll 15,000 individuals into its Autism Research Program and make de-identified genetic and other data available for research.
Snyder said that SensOmics differed from many of these other projects in that it seeks to go beyond genomics research and develop machine learning-based diagnostic tools.