Transitional disks are protoplanetary disks with gaps and holes. These gaps and holes are first discovered from the disks' SEDs (e.g. Spitzer), but later confirmed by direct imaging techniques. Recent near-IR polarization imaging (e.g. SEEDS, VLT) and submm (e.g. SMA, ALMA) observations have revealed a tremendous amount of details on these disks, but also posed new puzzles. These observations suggest that dust and gas start to decouple in these gaps and holes. The decoupling not only occurs in the inner disk but also in the outer disk, and moreover the decoupling can occur non-axisymmetrically in disks. To understand these new findings, we have carried out both hydrodynamic and magnetohydrodynamic global simulations with young planets in disks. Furthermore, dust particles have been implemented into these simulations to understand dust-gas interaction in turbulent disks. Non-ideal MHD effects have also been included in these global simulations. By comparing such realistic simulations with observations, we have constrained protoplanetary disk properties and revealed the early stage of planet formation. Finally, I will discuss the observational strategies to directly find young planets in protoplanetary disks and suggest that accreting circumplanetary disks could be the key to detect young planets directly. Current direct imaging observations have already found some circumplanetary disk candidates. In order to distinguish such disks with young planets, it is crucial to carry out observations at mid-IR.