Comparative assessment of methods for the computational inference of transcript isoform abundance from RNA-seq data