SPATA: A seeding and patching algorithm for de novo transcriptome assembly

RNA-seq reads are sampled from the underlying human transcriptome sequence, consisting of hundreds of thousands of mRNA transcripts. De novo transcriptome reconstruction from RNA-seq reads is a promising approach but facing algorithmic and computational challenges derived from nonlinear transcript structures and ultra high-throughput read counts. To tackle this issue, we designed a divide-and-conquer strategy to perform reads localization followed by a novel algorithm to assemble reads de novo. Using simulation studies, we have demonstrated a high accuracy in transcriptome structures reconstruction.