Label-Assemble: Leveraging Multiple Datasets with Partial Labels (preprint)

The success of deep learning relies heavily on large labeled datasets, but we often only have access to several small datasets associated with partial labels. To address this problem, we propose a new initiative, "Label-Assemble", that aims to unleash the full potential of partial labels from an assembly of public datasets. We discovered that learning from negative examples facilitates both computer-aided disease diagnosis and detection. This discovery will be particularly crucial in novel disease diagnosis, where positive examples are hard to collect, yet negative examples are relatively easier to assemble. For example, assembling existing labels from NIH ChestX-ray14 (available since 2017) significantly improves the accuracy of COVID-19 diagnosis from 96.3% to 99.3%. In addition to diagnosis, assembling labels can also improve disease detection, e.g., the detection of pancreatic ductal adenocarcinoma (PDAC) can greatly benefit from leveraging the labels of Cysts and PanNets (two other types of pancreatic abnormalities), increasing sensitivity from 52.1% to 84.0% while maintaining a high specificity of 98.0%.

Two Color Single Molecule Sequencing on GenoCare 1600 Platform to Facilitate Clinical Applications (preprint)

With the rapid development of precision medicine industry, DNA sequencing becomes increasingly important as a research and diagnosis tool. For clinical applications, medical professionals require a platform which is fast, easy to use, and presents clear information relevant to definitive diagnosis. We have developed a single molecule desktop sequencing platform, GenoCare 1600. Fast library preparation (without amplification) and simple instrument operation make it friendlier for clinical use. Here we presented sequencing data of E. coli sample from GenoCare 1600 with consensus accuracy reaches 99.99%. We also demonstrated sequencing of microbial mixtures and COVID-19 samples from throat swabs. Our data show accurate quantitation of microbial, sensitive identification of SARS-CoV-2 virus and detection of variants confirmed by Sanger sequencing.