The machine-learning-based method identifies relationships between bacterial strains and tracks their movements in less time, using less memory than existing solutions.
The solution uses machine learning to pare down somatic variant lists, simplifying the task of manually reviewing the output of sequence analysis pipelines.
The company has not released data on its method, but describes a circulating tumor DNA approach that has similarities to what other firms are pursuing.
Using the method, the researchers predicted the effects of over 140 million mutations in different tissues and identified mutations possibly associated with increased risk of several immune diseases.
MIT and Stanford computer scientists developed a technique for secure, massively scalable genomic analysis that they hope will unleash greater data sharing.
