MIG Seminar Series - Eduardo Eyras - Long read transcriptomics...beyond model organisms

Seminar/Forum

MIG Seminar Series - Eduardo Eyras - Long read transcriptomics...beyond model organisms

Long read transcriptomics - beyond model organisms

Abstract: New long-read sequencing technologies provide an unprecedented opportunity to study transcriptomes from any sample and species. However, analyses so far rely on a genome sequence or use hybrid approaches in combination with short reads, limiting the scope of the technology. The lack of dedicated tools hinders the full deployment of long-read technologies for the systematic and cost-effective interrogation of transcriptomes from any sample and species. This is crucial for non-model organisms, which are relevant to ecology, agriculture, and medicine, but also for model organisms, as available genome sequences do not always capture individual variations. We have developed a new method, RATTLE, for the reference-free reconstruction of transcriptomes from Oxford Nanopore cDNA/RNA sequencing reads. RATTLE uses similarity based clustering to reconstruct genes and alternative transcript isoforms, and to perform error correction from long reads without the need of a genome reference. Using experimental and simulated data and comparing to other methods, we show that RATTLE improves the mapping of long reads to the reference and delineates precisely exon-intron structures. RATTLE closes the existing technological gaps to build and interpret transcriptomes using long-read sequencing without relying on a genome sequence or additional technologies.

Presenter

  • Professor Eduardo Eyras
    Professor Eduardo Eyras, EMBL Group Leader and Professor