We present a brief review of the existing methods to estimate the dark matter content in the Universe, which are based on the strong lensing phenomenon, namely, the method of brightness ratio anomalies and the method based on the analysis of magnification probability distributions of macroimages caused by microlensing. Both of the methods need the macrolens model, the information on the source spatial structure, or at least on its effective size, and the following numerical simulation of microlensing events with subsequent comparison of the simulation results with observation data. Using the quadruply lensed quasar Q2237+0305 as an example, we show the effect of the source spatial structure on the shape of magnification probability distribution due to microlensing. A necessity to account for the contribution of the quasar intrinsic variability into the observed light curve is noted, as well as the need in developing a physically motivated algorithm to approximate the data of observations. With the use of all available observations of Q2237+0305, including our observations, the first histograms of microlensing amplification probability were built for all the four macroimages. We analyse the possibility to use them for evaluation of the dark matter content in the Q2237+0305 galaxy at the distance from the nucleus which corresponds to locations of macroimages.