Possibility to generate long RNA-seq reads such as Oxford Nanopore Reads can greatly improve transcriptome assembly and сonsequently complement the existing annotations of unfinished and poorly annotated genomes. De novo transcriptome reconstruction from short reads remains an open challenging problem, which is complicated by the varying expression levels across different genes, alternative splicing and paralogous genes. Indeed, long reads may be useful in resolving the later two problems and allow to generate more full-length isoforms. In this work we present computational methods for incorporating hybrid assembly into rnaSPAdes and provide a real life usage of this new pipeline by improving Artichoke gene annotation. We provide quality assessment reports for rnaSPAdes assemblies with and without nanopore reads, and also show impact to gene annotation caused by long reads.