Digital transcriptome subtraction or DTS was developed by Feng et al. to discover new viral agents associated with human cancers. Work on this approach was begun over a decade ago, prior to the release of the full human genome. It has only become a practical approach to new virus discovery with the generation of curated and highly-verified human genomic sequence databases and development of low-cost, high-throughput sequencing. This technique subtracts known human sequences from expression library data sets in silico, leaving candidate nonhuman sequences for further examination. Complementary DNA (cDNA) is created from RNA extracted from fresh tumor tissues through reverse transcription and then this library is exhaustively sequenced to generate a HiFi transcriptomedatabase of the tumor. Precise discrimination between human and non-human cDNA sequences is then performed and known human sequences that align to GenBank human reference sequences are then computationally subtracted. Careful screening and selection of high quality sequences prior to computational subtraction is required to increase chances of getting a true candidate. The non-matching non-human sequences that remain are candidate pathogen cDNAs. This smaller candidate database can then be explored using low-stringency alignment to virus databases or by experimental extension to identify novel viral sequences.
For more information about DTS for Short-Tag Sequences and High Throughput Sequences check out our DTS Tools page.