More Recent Comments

Thursday, April 20, 2017

The last molecular evolution exam: Question #4

More than 90% of our genome is transcribed when you add up all the transcripts from various cell types and various times of development (= pervasive transcription). Many biologists take this as evidence that most of the DNA in our genome is functional. What are the counter-arguments? Who do you believe and why?

Question #1, Question #2, Question #3, Question #4, Question #5, Question #6


3 comments :

Eric said...

A transcript is only evidence for transcriptase activity, not the activity of the transcript itself. Leaky RNA transcriptase activity can transcribe junk DNA and lead to non-functional RNA trascripts. Also, recently formed pseudogenes can produce a large number of non-functional RNA trascripts if the gene promoter is still operational.

Therefore, the null hypothesis is that transcripts are non-functional. You need evidence for function outside of the mere existence of an RNA transcript.

Anonymous said...

DNA has huge numbers of sequences that can function as promoters, and enzymes that transcribe DNA aren't all that specific. They'll even transcribe nonsense sequences that researchers make. Therefore, a low level of transcription is expected even at sites that lack promoters and don't produce functional proteins.

Low transcription levels don't in themselves prove that some sequences are junk (though they are suggestive), but they certainly don't prove that everything is functional.

judmarc said...

Much less than 90% of the genome shows selection (for or against). For those sequences that don't show selection, either sequence is not important, or variation is being "selected for:" that is, it's proving advantageous for survival of the population to have a variety of alleles.

So is variation being selected for? Since such amounts of variation (such a level of "junk") are found in relatively small populations where it can be accounted for by drift, there doesn't appear to be any need to posit "selection-for-variation" as a cause. In other words, drift would be the more parsimonious hypothesis.