The existence of multiple TSS for a gene is a key event to create diversity and flexibility in the regulation of gene expression under differential conditions (biotic or abiotic). The genomes of most eukaryotes are composed of transposable elements (TEs). TEs are also reported to be carrier of significant signals for the initiation of RNA synthesis and processing. Thus, the presence/absence of TE near the 5’ region of a gene may have a role in creation of new TSS leading to the expression of a new isoform. In this context we develop a pipeline to analyze RNA-seq data in the context of gene expression associated with structural changes (transcript isoforms) in different conditions, especially the study of hypothesis that TEs inserted in the vicinity of genes may affect structural changes of transcripts. In the last few years, a number of transcriptome assemblers have been developed, but the real challenge is to choose one of the existing assemblers that perform well enough for all data with different transcriptome complexities. We will present here preliminary results of our benchmarking to compare three different RNA assemblers, Cufflinks, Trinity and Grit. Then, we will also discuss about some cases studies of relation between TE and new TSS depending on the experimental condition tested.