pI distribution of E.coli cell extracts

[Bernard P. Murray, PhD]

In article <bpmurray*STUFFER*-1112981346130001 at mac-daddy.ucsf.edu>,
bpmurray*STUFFER*@socrates.ucsf.edu (Bernard P. Murray, PhD) wrote:

> In article <74qgqm$ens$1 at news.kren.nm.kr>, "Dae-Hyuk Kweon"
> <kdhbravo at plaza.snu.ac.kr> wrote:
> 
> > I think almost all proteins in E.coli have their own pI between 4-10.
> > But there will an exceptions.
> > I have 2D PAGE graph between 4-8.
> > I want to know the protein distribution curve of E. coli cell extract.
> > (pI versus density(regardless of kinds)).
> > What proportion of proteins of E.coli cell extracts will be exist above pI 8
> > ?
> > Thank you.
> 
> If you want a comprehensive but more theoretical answer
> and you have the computing resources you could predict
> the pI for each of the E.coli ORF protein sequences
> from the total K12 genome and work out the proportions
> from there.  The pI prediction (and it is just a prediction)
> is fairly trivial so all you need is the disk space.
> If I have room on my Linux box I'll give it a try over
> the weekend.

My apologies!  I underestimated the zeal of the E.coli people
in analysing their data.  They have already predicted the pI
values for all of the predicted 4290 protein-encoding ORFs.
I went to the Wisconsin K12 page at:
     http://www.genetics.wisc.edu/html/k12.html

and grabbed the Excel spreadsheet from:
      ftp://ftp.genetics.wisc.edu/pub/analysis/m52orfs.zip

Just to make things easier for you the pI values are grouped
as follows:

pI        Number of ORFs
==        ==============
<4        27
4 - 5     534
5 - 6     1057
6 - 7     875
7 - 8     358
8 - 9     573
9 - 10    638
10 - 11   178
11 - 12   44
>12       6

So *theoretically* about 33.5% of E.coli proteins will have
pI values greater than 8.  So, you should be reasonably
correct in your assumption that most pI values are 4 - 10
as only 5.4% lie outside this range.
     Bear in mind that many ORFs are only hypothetical,
post-translational modification is not accounted for, and
the calculated pI values do not always correspond well with
the actual values.
     I hope that this helps (I know I learnt something),
          Bernard
-- 
Bernard P. Murray, PhD
Dept. Cell. Mol. Pharmacol., UCSF, San Francisco, USA