How do you reduce backgrounds...?

[Andrew Cockburn]

In article <94143.084715PXH at psuvm.psu.edu>, Deepti Pradhan <PXH at psuvm.psu.edu> writes:
> I would really appreciate any suggestions from nettors on how I could reduce
> the nonspecific signals I am getting in my cDNA library screening.  I should
> mention that my hybridization and washes are at 68o [SSC protocol, no
> formamide here].  I did not have this problem before, when I was using an
> oligo probe.  The first time I saw this I thought my DNA was not clean , so
> I cut it again [it's in bluescript], extracted, made fresh probe and I
> *still* see everything light up.  Since it is a primary screen, it's got to
> be nonspecific signals.  BTW the library is in gt11.  This is the first time
> I am doing this stuff, so *any* [constructive :)] suggestions would be
> appreciated!
> Thanks,
> Deepti

There are two possibilities: non-specific binding and hybridization from
an unwanted contaminant (such as E. coli DNA).  You can distinguish these
by boiling the filters to see if the counts come off.  

If they do, then *something* real is hybridizing.  The most common problem
is E. coli DNA contaminating a plasmid prep, but it sounds like you should
not have this.  Both gt11 and bluescript have lac Z sequences in them, are
you sure that you are not getting vector to vector hybridization?  Finally,
you could have something like an Alu in your probe and be getting real
hybridization to every insert (depends on what your library is).

If the counts do not come off, the problem is just non-specific binding.
Use 7% SDS in your prehyb and hybridization, wash your filters at least
six times, buy new filters, and the problem should go away.

Andrew Cockburn
USDA