Ten percent of positively selected variants are within 7 Megabase of the human genome

Ten percent of SNPs that have been found to be positively selected as per the dbPSHP are within a 7 Megabase region on chr 2. This one region is very different from all other regions that have atleast 100 positively selected SNPs in that it has more than 1500 such variants.

1    154500000    167400000    121
10    100100000    113400000    108
12    52200000    62400000    125
14    59800000    70900000    117
15    62900000    79100000    121
17    50350000    64950000    125
18    63400000    67900000    135
2    62400000    75750000    131
2    95500000    101450000    111
2    131100000    137750000    1581
2    175250000    181150000    150
2    194600000    201750000    150
22    28200000    43700000    106
5    9800000    15600000    111
5    106700000    113850000    108
6    25050000    40450000    144
9    97850000    101000000    106
9    122600000    135200000    168

If so many positively selected variants are within such a short distance of each other, can they be considered independent events? Should methods that estimate selection coefficient's consider the cumulative value or just per variant estimates?