NIEHS Diagnostic-Prognostic RNAseq

Infectious disease outbreaks like Coronavirus Disease 2019 (COVID-19) can overwhelm healthcare systems when screening tools are lacking or scarce. In the face of an ongoing COVID-19 pandemic and with single-plex qPCR fluorometry inadequate for high throughput clinical diagnostics, the demand for testing exceeds the capacity, leading to limited availability, long queue times, backlogs in COVID-19 diagnoses, and delayed access to specialized treatment for COVID-19 patients. The goal of this RADx funded project was to develop an easily scalable and massively paralleled multiplexed screening method by next generation sequencing (NGS) with sample-specific barcoded indexes to allow pooling of up to 9200 samples. The approach enabled detection of SARS-COV-2 viral gRNA content (load) and expression profiles; detection of host's transcriptional response to infection simultaneously (expression profile). Transcriptional profiles were analyzed in the context of clinical data (electronic medical records of symptoms, treatments and outcomes) in order to identify and select features (biomarkers) that were predictive of clinical outcome using machine learning methods. These selected biomarkers were then validated in an independent clinical sample. Comparative testing of alternative sample collection (NP swab, saliva, oral swab) and handling protocols is ongoing with the goal of matching new protocols with existing SOPs for PCR-based testing in CLIA-certified facilities. In addition to increasing testing throughput, this precision medicine approach may have provided a very early look at who may be asymptomatic but infectious, or who is about to get very sick. Methods developed during this project may be adaptable to viruses or other pathogens in future pandemics. The initial aim of this project was to demonstrate proof of principle that the methodology can capture both virus and host RNA sequence using commercially available anonymous NP, OP and saliva samples.