I've been reading papers on RNA-Seq and it seems Illumina is the dominant technology at this point, mostly because the large number of reads. I am interested in 454 and want to know about coverage necessary for RNA-seq.
Some papers have shown that for mammalian genomes, the number of unique start sites reaches a plateau after about 80 million (non-unique) mapped reads for Illumina 35 bp length reads. Someone else has shown that 80% of yeast genes were detected after 4 million uniquely mapped reads (once again Illumina).
Obviously, you don't get near that many reads on 454 reads per run, but the reads are much longer. What are people's thoughts on the number of reads needed from 454 (say Titanium) to get acceptable coverage for RNA-seq for the human transcriptome. Any opinions out there?
Keep in mind, the goal here is not to necessary hit rare transcripts, but find differences in most common transcripts, e.g. tryi to mimic what is done with microarray 2 group comparison...
Thanks,
Wade
Some papers have shown that for mammalian genomes, the number of unique start sites reaches a plateau after about 80 million (non-unique) mapped reads for Illumina 35 bp length reads. Someone else has shown that 80% of yeast genes were detected after 4 million uniquely mapped reads (once again Illumina).
Obviously, you don't get near that many reads on 454 reads per run, but the reads are much longer. What are people's thoughts on the number of reads needed from 454 (say Titanium) to get acceptable coverage for RNA-seq for the human transcriptome. Any opinions out there?
Keep in mind, the goal here is not to necessary hit rare transcripts, but find differences in most common transcripts, e.g. tryi to mimic what is done with microarray 2 group comparison...
Thanks,
Wade