Virus identification has historically been performed through cell culture isolation and low-throughput methodologies that are limited often requiring previous information about the investigated viruses. These classical virological methods have been pivotal to many discoveries, but overall, they have a limited capacity for characterization of highly divergent and novel viruses. Nowadays, new technologies such as next-generation sequencing have revolutionized the virology field, enabling unbiased high-throughput viral genome characterization. But although the sequencing bottleneck has been surpassed, we could not say the same for the bioinformatics bottleneck of fishing new viral genomes from these large datasets littered with host and other microbes sequencing data. Here, we describe a bioinformatic framework to process metagenomic or metatranscriptomic data, aiming to assemble, identify, and study the evolutionary relationship of viral sequences and genomes.