A gene regulatory network (GRN) is a group of molecular regulators that operate together and with other components of the cell to control the amounts of mRNA and protein gene expression, which in turn controls the function of the cell. Single-cell gene expression can be quantified using scRNA-seq technology. Finding genes that co-express or show associated expression patterns between cells is made possible by these high-resolution data. GRNs can be constructed with the assistance of co-expression patterns, which may indicate possible links between genes. The reconstruction of GRNs is of cardinal importance in unraveling the complicated regulatory mechanisms in biological systems at the single-cell level. In this paper, we have used the GSE81252 dataset from NCBI-GEO related to liver tissue single-cell RNA-sequencing. We applied Long Short-Term Memory (LSTM), a deep learning method, for reconstructing GRNs. We obtained an accuracy of 94.74% with the LSTM model. These results point out the superior capability of LSTM networks in reconstructing GRNs from single-cell data. Thus, our results demonstrate that deep learning methodologies are feasible and have the potential to provide a better framework for future efforts on GRN reconstruction.