Presumption for the factor in binding discrepancy w/ in vitro values, same ligand etc

[Nagelfar]

One (in the same paper, Singh's Cocaine Antagonist one) IC50 value for cocaine has:

DAT w/ [3H]WIN 35428:
249 ± 37

SERT w/ [3H]paroxetine:
615 ± 120

NET w/ [3H]Nisoxetine
2500 ± 70

5-HTT/DAT selectivity
2.5
NET/DAT selectivity
10.0

The other is extremely different; also IC50 values (could it be a typographic error for Ki?):

DAT w/ [3H]WIN 35428:
89 ± 4.8

SERT w/ [3H]paroxetine:
1045 ± 89

NET w/ [3H]Nisoxetine
3298 ± 293
5-HTT/DAT selectivity
11.7
NET/DAT selectivity
37.0

There is no mention of different brain tissues (rat, monkey, etc.) used or other conditions. What could be the variable in these two studies on different tables from the same set of compounds to account for this difference? Even IC50 or Ki values for ligands from different labs seem to be within one point difference from one another; and this is the very same research paper (Page #973 versus #976 (49th versus 52th page of article) Table 28 VS. Table 29)

 

[serotonin2A]

One (in the same paper, Singh's Cocaine Antagonist one) IC50 value for cocaine has:

DAT w/ [3H]WIN 35428:
249 ± 37

SERT w/ [3H]paroxetine:
615 ± 120

NET w/ [3H]Nisoxetine
2500 ± 70

5-HTT/DAT selectivity
2.5
NET/DAT selectivity
10.0

The other is extremely different; also IC50 values (could it be a typographic error for Ki?):

DAT w/ [3H]WIN 35428:
89 ± 4.8

SERT w/ [3H]paroxetine:
1045 ± 89

NET w/ [3H]Nisoxetine
3298 ± 293
5-HTT/DAT selectivity
11.7
NET/DAT selectivity
37.0

There is no mention of different brain tissues (rat, monkey, etc.) used or other conditions. What could be the variable in these two studies on different tables from the same set of compounds to account for this difference? Even IC50 or Ki values for ligands from different labs seem to be within one point difference from one another; and this is the very same research paper (Page #973 versus #976 (49th versus 52th page of article) Table 28 VS. Table 29)

IC50 values are dependent on the experimental conditions. That's why it is preferable to convert the data to Ki. The idea with IC50 is that you can compare data obtained using the same experimental techniques (e.g., in a single study). But you can't necessarily compare different data sets.