PACIFIC

PACIFIC classifies SARS-CoV-2 and co-infecting respiratory RNA viruses from RNA-seq data using a deep-learning model to detect and quantify viral presence.

Key Features:

• Deep-learning algorithm: Uses a deep-learning classifier trained to recognize SARS-CoV-2 and common respiratory RNA viruses from RNA-seq datasets.
• High precision and recall: Achieves >99% precision and recall in detecting viral presence and quantifying relative concentrations on in silico data.
• End-to-end analysis: Performs end-to-end detection and quantification of viral sequences from RNA-seq inputs.
• Broad applicability: Validated across 63 independent datasets from in vitro cell culture experiments and patient samples.
• Lightweight implementation: Designed to operate without requiring extensive computational resources.

Scientific Applications:

• Monitoring viral infections: Systematic identification and quantification of SARS-CoV-2 and co-infecting respiratory RNA viruses from RNA-seq data.
• Co-infection dynamics: Analysis of co-infection patterns that may influence disease progression and patient outcomes.
• Epidemiological studies: Population-level surveillance of viral prevalence and co-infection frequency using sequencing datasets.
• Clinical diagnostics and therapeutic research: Support for diagnostic assessment and studies of therapeutic impacts related to viral co-infections.

Methodology:

The deep-learning model was trained on in silico RNA-seq datasets to detect and quantify SARS-CoV-2 and other respiratory RNA viruses.