Post Reply 
 
Thread Rating:
  • 0 Votes - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
A-site assignement in ribosome profiling experiments
09-Jan-2014, 10:28 AM
Post: #1
A-site assignement in ribosome profiling experiments
Hi, assuming that we have a Gaussian distribution of ribosome profiling read lengths centered around 30nts, how can we know exactly where the A site is for a particular read?
Find all posts by this user
Quote this message in a reply
09-Jan-2014, 02:16 PM
Post: #2
RE: A-site assignement in ribosome profiling experiments
Hi Antón,

That is a very interesting question because for a given ribosome protected fragment, how to ascertain if the variation in read length originates from the 5' or the 3' end of the read?

In GWIPS-viz, for the eukaryotic ribo-seq data, we typically use a 15nt off-set from the 5'end of the RPF to the A-site , unless the study publication states otherwise. However, this is an average. Some studies adjusted the offset according to read-length.

For prokaryotic ribo-seq data, we use the weighted centered approach implemented by (Oh et al., 2011) to indicate the putative location of the decoding centre.

It is a very pertinent question and if other users have comments or suggestions, it would be great to hear them.

Thanks,
Audrey
Find all posts by this user
Quote this message in a reply
09-Jan-2014, 02:21 PM
Post: #3
RE: A-site assignement in ribosome profiling experiments
Hi Anton, welcome to the forum!

The truth is that no one knows what is the distribution of lengths of the reads protected by the ribosome. The technique involves a step where you cut fragments from the gel. This is a bit arbitrary, you could cut a thick slice, you could cut a very thin slice. I've heard some people advocating the latter because it increases triplet periodicity (I think this is both a very good idea and a very bad idea depending on what you are going to do with the data).

However, regarding the location of the A-site. Lets assume that there is a minimal length of a mRNA fragment that is protected by the ribosome (essentially the length of ribosomal mRNA tunnel measured in nucleotides using well stretched RNA molecules as a ruler), say it is x=30 nucleotides. Because the ribosome is more or less rigid, the A-site position would be more or less fixed relative to the ends of the tunnel and to the ends of protected reads. The different story is when you have a read whose length is higher than x. In this case you actually do not know where the A-site is because you do not know whether the ribosome sucked up more RNA at the 5' or the 3' end. Because bacterial ribosomes tend to protect longer fragments, you cannot set an offset to predict the A-site location, hence a center-weighted approach has been used and we adopted this approach for GWIPS browser for bacterial data.

However, clearly inaccuracies are expected with such approach and the distribution of the reads is unlikely to be Gaussian. See http://www.ncbi.nlm.nih.gov/pubmed/23603333 for some reasons behind that. Also I recently came across of a report of bimodal read length distribution here: http://www.ncbi.nlm.nih.gov/pubmed/24301020

primus inter pares Angel
Find all posts by this user
Quote this message in a reply
21-Jan-2014, 05:33 PM
Post: #4
RE: A-site assignement in ribosome profiling experiments
It's not that bacterial ribosomes tend to protect longer fragments, it's that the nuclease used for bacterial ribosome profiling has biases and will not completely digest unprotected mRNAs. One could write a more sophisticated program to account for these biases, but due to this bias, codon resolution for bacterial ribosome profiling will necessarily take a hit. A center-weighted approach is likely to be good enough for most purposes.

For my own P-site assignment of reads (I chose to assign P site rather than A site), I split the reads into different read lengths and determined where is saw the codon periodicity / phasing - between read lengths, this phasing should align. This would represent the maximum-likelihood estimate of where the P-site would be. I've adapted my approach from Nick Ingolia's own approach (in his 2011 mESC paper).
Find all posts by this user
Quote this message in a reply
21-Jan-2014, 11:56 PM
Post: #5
RE: A-site assignement in ribosome profiling experiments
You mean that micrococcal nuclease does not digest mRNA completely? That's true, but only a part of the story, we wouldn't observe correlation between Shine-Dalgarno sequences and the length of the protected fragments if it was the only reason.

primus inter pares Angel
Find all posts by this user
Quote this message in a reply
04-May-2016, 08:23 PM
Post: #6
RE: A-site assignement in ribosome profiling experiments
Thanks for this post, this is very helpful. A question that comes to mind along these lines-- what was your decision point to use the P-site vs. the A-site? Also, when doing your analyses, do you keep fragments sizes between i.e. 28-33? What would be recommended for a human dataset?

Many Thanks,

(21-Jan-2014 05:33 PM)chewgl Wrote:  It's not that bacterial ribosomes tend to protect longer fragments, it's that the nuclease used for bacterial ribosome profiling has biases and will not completely digest unprotected mRNAs. One could write a more sophisticated program to account for these biases, but due to this bias, codon resolution for bacterial ribosome profiling will necessarily take a hit. A center-weighted approach is likely to be good enough for most purposes.

For my own P-site assignment of reads (I chose to assign P site rather than A site), I split the reads into different read lengths and determined where is saw the codon periodicity / phasing - between read lengths, this phasing should align. This would represent the maximum-likelihood estimate of where the P-site would be. I've adapted my approach from Nick Ingolia's own approach (in his 2011 mESC paper).
Find all posts by this user
Quote this message in a reply
Post Reply 


Forum Jump:


User(s) browsing this thread: 1 Guest(s)

GWIPS-viz | Return to Top | Return to Content | Light (Archive) Mode | RSS Syndication