not enough isogroups in 454 transcriptome?

[litali]

hi,
we have sequenced via 454 a novel transcriptome for an organism with `~20000 genes, and received only 2000 isotigs/isogroups though there were enough reads and most of the reads were aligned to those 2000 isotigs. What could be the reason for so small number of isotigs/isogroups?

[flxlex]

How many reads do you have? It's either that, or you picked a sample/samples where only about 2000 genes are expressed...

[flxlex]

How many reads do you have? It's either that, or you picked a sample/samples where only about 2000 genes are expressed...

[litali]

reply-not enough transcripts

There are about 100000 reads assembled, but only 1000-2000 isogroups..can't it be something in the assembly level? It is very strange that only such low number of transcripts is expressed (even in the control sample).
thanks for your help!

[kmcarr]

What was the source of your RNA sample, as in what species and was it a specific tissue, organ or cell type?

[genlyai]

It may have more to do with the distribution of transcript expression than the number of genes "expressed". If you sequenced whole blood, so much of your transcript would be hemoglobin that you would see very little of anything else. This would push even moderately expressed genes to very low read numbers.

Also, even in a tissue where one or a few transcripts do not dominate, the number of transcripts per cell is usually said to be about 350,000. This means that with 100,000 reads, anything expressed at <10 copies/cell -- in other words, the majority of expressed genes -- is going to have <3 reads & may not assemble. [Though I'm not sure if you have 100K starting reads or 100K assembled reads.]

[Jeremy]

From almost 500k reads of RNA-seq on 454 I got around 15k isotigs, so what you have found seems about right.

If you want to find the type of genes present and not worry about their expression levels you should try a normalized library.