sequencing the Exome of a new vertebrate

[Dan_cham]

hello to all

I want to sequence the Exome of a certain vertebrate with unknown genome (using SOLID).

I am going to extract RNA form tissue.

Does anyone have recommendation about which tissue (or tissues) to extract the RNA from, in order to get sequences that represent the Exome of the animal in the best way possible (in contrast to a tissue specific Exome - that might cause me to loose transcripts)

Thanks so much

Dan

[krobison]

Just a note on terminology: if you are sequencing mRNA, that is transcriptome sequencing.

Exome sequencing refers to targeting the exonic DNA of a creature.

Unless you homogenize the whole beast, you'll have some tissue specificity. Pooling equal amounts of mRNA from different tissues might be one way around, but you can't expect to get everything.

[KevinLam]

and if your aim is to get the entire cds of the organism,
do note that the RNA transcripts will be in unequal proportions so you will lose the rare transcripts unless u have high coverage, or am able to normalize the transcripts by identity ( if u can do the latter do let me know. I am keen to see how this is possible)

[Dan_cham]

I am aware of the problem of rare transcript and hope to deal with that using rRNA reduction kits and Poly A enrichment kit.

I will be glad to hear more suggestion and will let u know how it works (hopefully)

Thanks
Dan

[pmiguel]

Hi Dan,

With sequencing there are two different worlds. The "de novo" world and the "re-sequencing" world.

At this point, I don't know that I would recommend the SOLiD for de novo transcriptomes. The short reads and complications involved in dealing with color space reads would make non-reference sequence guided assembly a challenge. This will all change with ECC reads that are going to be possible on the SOLiD 5500.

On the other hand, I guess I have never tried assembling transcriptome reads generated by the SOLiD. Maybe it would work.

--
Phillip

[Dan_cham]

hi philip

I hope that because I an dealing with transciptom and not genome, the assembly problems will be... lets say - less difficult.

because of the fact that there aren't a lot of Repetetive elemnts.

my goal is to extract many genes (about 80) for a certain analysis. if i would be able to read the transcriptom it will be a bonus to me (but it is not the main goal)

do u know when the 5500 will be available and what make it so unique?

Thanks

[pmiguel]

The issue is that the color space data generated by the SOLiD is ideal for certain applications, eg. SNP discovery, but less so for other, eg de novo assembly. Only a few assemblers understand color space natively. Color space reads can be converted to base space, But in doing so, not only do you lose the error detection built into color space, but you ruin a large fraction of your reads because of a certain type of error propagation in the conversion process.

Probably most sites that plan to upgrade to the SOLiD 5500XL will have done so by summer. One of the 5500's capabilities is that it can run the new "ECC", "Exact Call Chemistry". This is another set of ligations that solves the error propagation issue with non-ECC reads. Hence conversion to base space works fine on these types of SOLiD reads. Once in base space most assemblers will work on these reads.

Anyway, should be possible to do what you want in any case. But if I were looking at your project, I would think "454" or "Illumina", not "SOLiD". Not that SOLiD data cannot be used in the way you intend, just that the down stream infomatics will be more challenging.

--
Phillip

[Dan_cham]

Thanks Phillip

I know that 454 or ilumina is better for de novo but I hope that because i will use end paired it will help me in my goal.

Dan

[krobison]

Paired end will help (though beware! in the fall the "reverse" reads were not fully debugged) but not with any of the issues raised above -- there is a shortage of color-space aware aligners & if you convert naively from colorspace to basespace there is the colorspace equivalent of a frameshift lurking (decoding one color incorrectly shifts the interpretation of those downstream)

[Joann]

Embryology

You might want to work with an embryo(s) at a characteristic stage of development. That would damp down the relative constitutive content and most likely enhance for species relevant expression. Plus the data you obtain would be immediately relevant to a comparison study of another developmental stage.