I've downloaded a multiple sequence alignment of 100 species for a fragment of human ACTB gene (specifically for one of it's exons; coordinates are chr7:5567961-5568345) from UCSC table browser. The file itself is available at https://www.dropbox.com/s/tvm35yydtjn7jvx/ACTB.some_exon.maf. When looking at this region in UCSC genome browser, it seems to be conservative: However, when I run PhyloP on this alignment, p-value of conservation is 1:

$ wget -q -O ACTB.some_exon.maf 'https://www.dropbox.com/s/tvm35yydtjn7jvx/ACTB.some_exon.maf'
$ wget -q 'http://hgdownload.cse.ucsc.edu/goldenpath/hg19/phyloP100way/hg19.100way.phyloP100way.mod'
$ phyloP hg19.100way.phyloP100way.mod ACTB.some_exon.maf
WARNING: pruned away leaves with no match in alignment (papHam1, pteVam1, echTel2, dasNov3).

*****
P-values for number of substitutions observed in 'ACTB.some_exon.maf' given 'hg19.100way.phyloP100way.mod'
*****

p-value of conservation: 1.000000e+00
p-value of acceleration: 0.000000e+00

null distrib: mean = 883.012452, var = 910.568347, 95% c.i. = [824, 943]
posterior distrib: mean = 4554.053300, var = 4839.759760

Am I running PhyloP incorrectly?