Using two novel methods, pair invariant mass (minv) and comparative measurements with respect to reaction plane (ΨRP) and participant plane (ΨPP), we isolate the possible chiral magnetic effect (CME) from backgrounds in 200 GeV Au+Au collisions at STAR. The invariant mass method identifies the resonance background contributions, coupled with the elliptic flow (ν2), to the charge correlator CME observable (Δγ). At high mass (minv>1.5GeV/c2) where resonance contribution is small, we obtain the average Δγ magnitude. In the low mass region (minv<1.5GeV/c2), resonance peaks are observed in Δγ(minv). An event shape engineering (ESE) method is used to model the background shape in minv to extract the potential CME signal at low minv. In the comparative method, the ΨRP is assessed by spectator neutrons measured by the ZDCs, and the ΨPP by the 2nd-harmonic event plane measured by the TPC. The ν2 is stronger along ΨPP and weaker along ΨRP; in contrast, the magnetic field, mainly from spectator protons, is weaker along ΨPP and stronger along ΨRP. As a result, the Δγ measured with respect to ΨRP and ΨPP contain different amounts of CME and background, and can thus determine these two contributions. It is found that the possible CME signals with background isolation by these two novel methods are small, on the order of a few percent of the inclusive Δγ measurements.