Central Dogma (What use cDNA)?

[Simon Penson]

Dear Netters,

I've a few questions about the logic behind homology-based approaches to
`finding' proteins. I'd be grateful for any non-smartarsed replies!

1. Although it's hugely variable between species, is there a ball-park
figure for the % of DNA in a genome which is known to code for
functionally-expressed proteins? I've heard as low as 1%, but much higher
for Arabibopsis.

2. Of all the mRNA which is transcribed, what % is actually translated and
prossesed into functional protein?

What I'm driving at is: if you pull out a cDNA clone and do a homology
search, what degree of certainty do you have that the putative protein
ever exists in the cell?

Cheers,

Simon.