Experiment Bank - The site for free exchange, review, and discussion of novel scientific experiments Front Page ? Everything ? News ? Wish list ? Help ? FAQ
?Alternative Splicing QTL

cellular and molecular biology
By gbaute, Section Biology
Posted on Fri Apr 30th, 2010 at 11:54:02 PM PST

Alternative splicing (AS) is a essential form of gene regulation effecting up to 95% of genes in some eukaryotes, key players in its control are poorly understood. A large component of this unknown is the contribution of cis (signals on the RNA) and trans (mostly protein machinery) regulation. This experiment would identify the contribution of both as well as which regions in the genome are responsible.

This experimental layout could be used for any species with AS but I think drosophila, which has very prevalent AS would be a good choice. Two divergent strains with different AS patterns will be crossed and then back-crossed to make a QTL population. Illuminia runs for each line (depending on funding could be 25-250 lines) will essentially do both parts of the experiment at once, it will assess transcriptome wide AS patterns and find (by allele specific snps) the contribution of each parent to the genome. QTL analysis will reveal which AS events are associated with which genomic regions. This is very similar to eQTL experiments. If most AS events are associated with the loci that the genes come from then cis control is important for that allele if its from another region then its trans.

This would further our understanding of AS and possibly reveal new mechanisms of its control. A bonus will be assessing how AS may have diverged in these species.

 

< Alternate Interactomes (1 comments) | Genome methylation and folate intake (4 comments) >



View: Display: Sort:
Alternative Splicing QTL | 6 comments (4 topical, 2 editorial)
[new] Natural variation of expression (none / 0) (#1)
by xerro five on Wed Apr 28th, 2010 at 11:54:38 AM PST
(User Info)

This is a good idea, but I have a concern about determining the amount of each AS isoform. This is because the level of expression of a given isoform can vary due to natural and uncontrollable events (and which are presumably not differences stemming from the sequence elements you are looking for) meaning that even between two genetically-identical organisms there will be natural differences between the levels. The real problem is that no one knows how much variation there is/can be. So then would this make noise to much of a problem? That is, do you expect the changes you are looking for to be large enough for them to stand out against this background?



[new] AS level variance? (none / 0) (#2)
by schun on Fri Apr 30th, 2010 at 12:57:53 PM PST
(User Info)

I thought that between genetically-identical organisms, the isoform variance shouldn't be that big. For instance, if isoform1 is 90% and isoform is 10% in organism A, it should be for organism B as well so it shouldn't cause that much noise in that sense.

[ Parent ]


[new] thats right. (none / 0) (#5)
by gbaute on Sat May 1st, 2010 at 10:11:43 AM PST
(User Info)

AS is pretty highly conserved with in species.

[ Parent ]


[new] Schools of thought on AS (none / 0) (#4)
by gbaute on Sat May 1st, 2010 at 10:08:45 AM PST
(User Info)

Many people are of the opinion that AS is noise, falling into 2 categories, 1)technical and 2)biological (miss splicing). Technical stuff is a problem yes. This experiment is setting out to exactly pin down what is causing (namely cis or trans factors) the different splicing forms to exist in different organisms. Genetically identical individuals will have the same AS patterns if grown in the same conditions. More dissimilar organisms will have different AS and this project could find out why.

[ Parent ]


Interesting proposal none (#3)
by timothy auyeung on Fri Apr 30th, 2010 at 03:54:21 PM PST
(User Info)

This is a very interesting proposal. I wonder whether in the end we will simply find that most of the regulatory mechanisms governing AS will consist of both cis and trans components. This is because I presume that for most trans-acting factors, target specificity demands that they need a recognition site on the target of interest such that trans mechanisms are accompanied by cis elements.



yes. none (#6)
by gbaute on Sat May 1st, 2010 at 10:13:21 AM PST
(User Info)

I would expect the majority to be both cis and trans. It is possible that no new or clear picture emerges but you would never know until you tried.

[ Parent ]


Alternative Splicing QTL | 6 comments (4 topical, 2 editorial)
View: Display: Sort:
?+?Menu
? create account
? faq
? search

?+?Login
Make a new account
Username:
Password:

?+?Related Links
? More on cellular and molecular biology
? Also by gbaute

BC Cancer Agency Genome Sequence Centre SourceForge Logo Powered by Scoop
All trademarks and copyrights on this page are owned by their respective companies. Comments are owned by the Poster. The Rest ? 1999 The Management

create account | faq | search