Transcriptome Assembly: Number Of Transcripts Expected?

Question

[Rm]

I just completed a transcriptome assembly (~ 50 Million paired end 100bps read sets) of Xenopus species without reference genome:

I followed velvet + oases multi-kmer approach with min coverage 5 and and merged the transcripts obtained from multiple Kmer (31 to 75 with step of 6) assembles: In the merged set I got around 200K Loci with around 700K total transcripts?

I was just wondering if these numbers in generally expected range? or high

Answer 1

[Istvan Albert]

The number of expected transcripts is the number of existing transcripts. You can estimate that from other related genomes because even though the genomes might vary wildly the number of transcripts will probably vary a lot less.

Of course it is unlikely that any method would ever be able to match that correctly and usually you will get a lot more.

[Rm]

@Istvan : Thank you; i will look over numbers from related species....

[Rm]

i used the CD-hit-est on the merged assembly at -c 0.9 it reduced the number to ~250K