Optimization of the Illumina COVIDSeq™ protocol for decentralized, cost-effective genomic surveillance.

A decentralized surveillance system to identify local outbreaks and monitor SARS-CoV-2 Variants of Concern is one of the primary strategies for the pandemic's containment. Although next-generation sequencing (NGS) is a gold standard for genomic surveillance and variant discovery, the technology is still cost-prohibitive for decentralized sequencing, particularly in small independent labs with limited resources. We have optimized the Illumina COVIDSeq™ protocol for the Illumina MiniSeq instrument to reduce cost without compromising accuracy. We slashed the library preparation cost by half by using 50% of recommended reagents at each step and normalizing the libraries before pooling to achieve uniform coverage. Reagent-only cost (∼$43.27/sample) for SARS-CoV-2 variant analysis with this normalized input protocol on MiniSeq instruments is comparable to what is achieved on high throughput instruments such as NextSeq and NovaSeq. Using this modified protocol, we tested 153 clinical samples, and 90% of genomic coverage was achieved for 142/153 samples analyzed in this study. The lineage was correctly assigned to all samples (152/153) except for one. This modified protocol can help laboratories with constrained resources to contribute in decentralized COVID-19 surveillance in the post-vaccination era.

Multi-Class Flat Classification of Lung Diseases Utilizing Deep Learning

Early diagnosis of diseases is very critical for recovery. However, this is not always feasible due to the limited available staff or expensive and inadequate tools as we have witnessed in the recent COVID-19 pandemic. Lung diseases are life-threatening, but fortunately, they can be detected from X-ray images, which are cost-effective approaches. However, they need experts who are sometimes unavailable. Thus, using cutting-edge technology to diagnose diseases automatically and fast is the key solution to saving people's lives. In this research, deep learning techniques have been utilized to classify several lung diseases in a cost-saving, time-saving, and efficient manner. Examples of lung diseases studied in this research are COVID-19, Lung Opacity, Pneumonia, and Tuberculosis. Several pre-trained deep learning models have been employed for flat multi-class classification of these lung diseases instead of using binary classification to recognize one disease from normal cases, as most state-of-the-art studies carry out. The models' performance has been evaluated on imbalanced data of X-ray images with various resolutions and types. Finally, multiple measurements metrics have been utilized to evaluate the performance. The best accuracy achieved in this research is 95.643%. © 2022 IEEE.