(by hzhen from freeuk.com)
Tue Feb 12 15:42:38 EST 2008
> In article <foounk$dd9$1 from oravannahka.helsinki.fi>, kaj.stenberg from helsinki.fi.invalid wrote:
>> DK <dk from no.email.thankstospam.net> wrote:
>>> In article <fomluv$bge$3 from oravannahka.helsinki.fi>,
>> kaj.stenberg from helsinki.fi.invalid wrote:
>>>> This is probably trivial, but it confuces me:
>>>> If the Kd for a ligand is mM it binds poorly, if the Kd is nM it binds
>>>> well. But in which case does one refer to the Kd as being "higher" or
>>> mM is "higher" than nM. However, sometimes people switch from
>>> talking about Kd to talking about "affinity". Affinity is "higher" when
>>> Kd ~ nM.
Yes, people tend to say "X has a high affinity binding with a Kd of
...", the reason probably being that affinity is easily understood
semantically, while the number from Kd is easily understood
scientifically. No problem if you know what affinity is, but many
>> Thank you, "affinity" is of course unambigious, I did just not come to
>> think of it.
>> Just another thing: when can you start talking about a "real" (as in
>> specific) affinity? I am reading papers with a Kd in the ballpark of
>> tens of millimolar and find it hard to repeat the work so that I truly
>> believe in the results, since every random event seems to affect the
>> curves fairly much.
> AFAIK, for proteins > mM affinities are usually derived indirectly
> from some other data (such as Km) rather than measured.
> There can never be a strict threshold but for protein-protein
> interaction the fuzzy boundary between real and not, specific
> and not lies somewhere around tens of microM Kd.
The bane of my life, trying to find out whether the binding of a protein
or peptide is real or has any significance. BTW, I had done experiments
where I showed that a binding in the low mM range is real and specific
(by NMR - you can show specific binding by the effect on particular
cluster of residues). Whether it has any significance is another matter
More information about the Methods