Would anyone tell me how I should assemble lowly expressed transcripts (which might be discarded by cufflinks) based on NGS-seq data? Can I do it myself...