Droplet Barcoding Test for Prenatal Diagnosis and MSI Cancer Analysis 

Rare DNA molecules are very difficult to identify but they usually play a very important role in many diagnoses. For example, the fetus's DNA only takes up a very low percentage in a pregnant mother’s blood. However, if we can identify these few DNA molecules, we can diagnosis the gene-based disease and monitor the healthy state of the baby in a very simple way. In the very early stage of cancer, tumor shed DNA into the bloodstream but may contain as little as 0.01% compared to the normal DNA. If we can identify this very little fraction of tumor DNA, which offers the potential for non-invasive companion diagnostic tests to guide cancer therapies and sequentially increase the survival rates of the patient. But, why is this rare DNA molecule detection so difficult? The reason is that we need a pre-amplification process prior to the detection. This amplification uses DNA polymerase such as Taq which are error-prone polymerase. These artifacts have, to date, prevented the detection of specificity of the target molecules from the baby or tumor. So we need a high fidelity method to amplify this rare molecule when keeping the specificity.

Our technology encapsulates single DNA templates into drops, amplifies them by PCR, and then labels them with “barcodes” prior to sequencing them via massively-parallel sequencing platforms, such as the Illumina MiSeq. The computational fusion of the PCR product sequences yields an error-free sequence of each input DNA template. The process uses microfluidic technologies to synthesize the drops, encapsulate DNA, introduce reagents into drops, and prepare the amplified targeted molecules for sequencing. We couple this sample preparation with software algorithms, and a graphical display to support all data quality analysis. We hope this technology will benefit not only scientific research but also clinical diagnosis areas such as cancer and prenatal diagnosis.

Contact: Ye Tao ytao@seas.harvard.edu