Sequencing is a fundamental tool for profiling and analysing microbiomes. Historically, sequencing-based microbiome research focused on prokaryotes, often overlooking viruses. To study the viral component, researchers typically used shotgun metagenomics, in which all DNA in a sample is randomly fragmented and sequenced. Metagenomes were assembled and viruses identified. However, assembly-based approaches are computationally intensive and have limited sensitivity, particularly for low-abundance taxa such as viruses. Virus enrichment techniques (prior to sequencing) can improve sensitivity but often result in the loss of information about other domains of life, such as prokaryotes. This is important because most viruses are phages — viruses that infect bacteria. Although viral enrichment can be used to gain insights into phages, it fails to capture data on their bacterial hosts. To overcome these limitations, we developed Phanta1, which is an optimized read classifier that enables rapid and sensitive profiling of viruses and prokaryotes from standard shotgun metagenomics data, providing a more comprehensive microbiome view.

When designing Phanta, our aim was to create a tool that can leverage the unbiased nature of shotgun metagenomics to quantify taxonomy across all domains of life. In addition, we aimed to develop a tool that would be sensitive enough to detect low-abundant taxa. We also prioritized speed and ease of use. By reducing the activation energy to perform phage-inclusive profiling, we aspire to make it the standard in the field.