Galaxies can form in a sufficiently deep gravitational potential so that efficient gas cooling occurs. We estimate that such potential is provided by a halo of mass $M gtsim M_{c} approx 7.0 imes 10^{12} ~ (Delta_{c}(z) (1+z)^{3})^{-1/2} Msun$, where $Delta_{c}(z)$ is the mean overdensity of spherically virialized objects formed at redshift $z$, and $M_{c} approx 4.0 imes 10^{11} Msun$ at $z = 0$. Based on this criterion, our galaxy samples are constructed from cosmology simulation data by using HiFOF to select subhalos in those FOF halos that are more massive than $M_{c}$. There are far more dark subhalos than galaxy-hosting subhalos. Several tests against observations have been performed to examine our galaxy samples, including: (1) The differential galaxy mass functions of this sample are found to be close to the observation derived from the combination of the luminosity function of the DEEP2 galaxies as well as the mass-to-light ratio from Red-Sequence Cluster Survey at $z = 0.3$. (2) The galaxy space density is analyzed as a function of redshift to examine the density evolution, which is found to be roughly consistent with the observational result based on the luminosity functions in cite{fab07}. (3) The projected two point correlation functions (CF) of our galaxy sample at $z$ = 0 and $z$ = 1 are in good agreement with those of the SDSS and DEEP2 galaxies, respectively. (4) The HODs of our galaxy samples show good agreement with the SDSS and DEEP2 data. (5) Finally, the kinematic pair fractions for $r_{max}$ = 50 and 100 $kpc$ are computed, and the evolution is parameterized as $(1+z)^{m}$. Comparing with the results, m = $0.41pm0.14$ ($r_{max}$ = 50) and m = $0.29pm0.05$ ($r_{max}$ = 100), in cite{lin08}, we found m $approx$ 0.96 and m $approx$ 0.48 respectively from one of our simulations. Based on the consistency with observations, our galaxy sample is believed to correctly represent galaxies in real universe, and can be used to study other unexplored galaxy properties.