Cell migration is a key step for deterioration of many in situ or metastasis malignant tumours. Tumour anti-migration is a promising strategy to treat cancer, but corresponding drugs developed under such a strategy are still in dire poverty, partly due to the lengthly process of drug trials and approval required by the US Food and Drug Administration (FDA). Given there are thousands of FDA approved drugs in the market, we believe that drug repositioning may provide a fast and cost-effective way to identify potential anti-migration drugs. In this paper, an in-silico drug screening method using a genomic strategy is proposed for the goal, in which genomic signature identification combined with support vector machine modelling is adopted to estimate drug efficacy. And a high-throughput, sensitive, 3-dimensional invasion assay by quantitative bioluminescence imaging proved the performance of proposed method on in vitro disease models.