Re: Alternative Splicing


Larry Moran wrote:
> On 30 Nov 2005 15:29:33 -0800, wade <wade.hines@xxxxxxx> wrote:
>
> [snip]
>
> > Clearly, if 10 exons for proteinX are conserved for all studied
> > vertibrate genomes, suggesting that a splice variant is mostly
> > unique to one species is dubious. But I do forsee some wonderfull
> > papers that I hope will soon be published "Most variable splice
> > claims are bunk". Wish I could take a few months off to do the
> > work and write it.
>
> I'm working on it. I've assembled all of the alternative splice data
> for human and mouse HSP70 genes. These genes are among the most
> highly conserved genes known. The predicted splice variants are
> supposed to produce an abundant array of truncated proteins with
> various inserts and deletions within the hydrophibic core and around
> the active sites of the protein.
>
> It makes no sense whatsoever that mammalian genomes would produce
> such strange variants of such highly conserved genes. The HSP70
> proteins play an important role as essential chaperones in all
> species. Furthermore, the mouse and human alternatively spiced
> variants are different. This is silly. It's not possible that
> humans and mice could have evolved strange variants of these
> genes in the past 100 million years when their structure was
> conserved for the previous 3,500 million years.
>
> The patterns of alternative splice variants are predicted from the
> EST data. The genome annotators have ignored this data for the HSP70
> genes because they know it's an artifact. The patterns are no
> different than those for all the other genes.
>
> Here's an example for HSPA5 (BiP/GRP78).
>
> http://spliceinfo.mbc.nctu.edu.tw/ps_splice_view.php?gs_id=ENSG00000044574
>
> This leaves us with three possibilities ....
>
> 1. I've made a serious mistake and the human/mouse HSP70 genes
> really do produce an abundance of alternatively spliced variants.
>
> 2. The HSP70 genes are very special. The EST data is no good for
> those genes but it's perfectly valid for all those other genes
> with mostly unknown protein products whose structures haven't
> been solved.
>
> 3. The entire set of EST data is flawed and other genes are just
> like the HSP70 genes. They don't have alternatively spliced
> variants.
>
> Wanna help with the paper? :-)

There's wanna and there's reality.

I think the thing to turn what you've started into a great paper would
be to have some solid examples of some proteins that are products
of variably spliced genes (like where there's legit data for
functioning
translation products made of actual amino acids). Observations on
how real proteins are conserved across species may be lost on the
vsplice sects but they might resonate with those who occupy an orbit
somewhat closer to Earth.

I have considered mining some of the proteomics data where there
are putative MS/MS spectral matches to peptides matching vsplice
forms of proteins. But then, I have other things to do that I actually
get paid to do and publications to write on things I've actually done.
Still, it might help settle a bet.

.