The antenna selection technology has become essential for massive multiple-input multiple-output (MIMO) systems since it allows us to benefit from the massive increase in the number of antennas without a necessary use of additional RF components. However, reducing the number of active antennas may exhibit a performance loss regardless of the used antenna selection technique. In this paper, we propose a low-complexity and near maximum-likelihood (ML) performance achieving receiver for large MIMO systems with antenna selection. The proposed approach is based on the likelihood ascent search (LAS) algorithm and allows to increase the system performance in terms of bit error rate (BER). The performance of the proposed detector is analyzed by Monte-Carlo simulations. Numerical results show that our technique provides a gain of about 1 dB.