Dear all:
I hope I have post it on the right place. I have been studying
polymorphisms of HLA-DRB1 genes and would like to ask some questions
I can't solve by myself. Here is the problem:
I found from the exon2 sequences that, residues 11 and 13 of
HLA-DRB1 molecules are corelated with serologically typing result.
Please see the following table. ("serotypes" is the first two numbers
after the asterisk of the nomemclature of DRB1 genes. for example
"HLA-DRB1*0101" -> serotype 01 )
================================
residue 11 | serotypes
-------------+------------------
L | 01
P | 15 16
S | 03 11 12 13 14 08
V | 04 10
G | 07
D | 09
================================
With two exceptions DRB1*1410 DRB1*1211
residue 10 and 12 do not display such strong correlation with serotypes:
================================
residue 10 | serotypes
-------------+------------------
Q | 01 16 04 07 09
Y | 03 11 12 13 14 08
E | 10
================================
I even perform a chi-sqare test on sequences data, the correlation score is
===================
residue | score
---------+---------
9 | 1.414214
10 | 1.404567
11 | 2.176344
12 | 0.981129
13 | 2.159943
===================
So, I have thought that, residue 11 and 13 must be polymorphic antibody
epitopes or they can change the conformation of other epitopes so as to
influence antibody definitions. However, I read from a paper that
substitution at residue 10 and 12, but not 9, 11, 13 can affect antibody
recognition, which seems to be a reverse of my statistical predictions.
Do you know why?
Greatly appreciate your help in advance!
yours sincerely
feng, Bingjian
