Abstract: The present discloses a predicting method of transcription factor binding sites based on weighted multi-granularity scanning, which belongs to the field of site prediction. The method comprises: using an inverse sequence, a complementary sequence and a complementary inverse sequence to augment an initial data set; carrying out feature representation on the DNA sequence by combining one-hot coding and multi-base feature coding; dividing a training set and a test set; calculating weight vectors of the features; carrying out weighted multi-granularity scanning; performing model training through a cascade forest to obtain a classification prediction model of transcription factor binding sites; inputting the test set into the classification prediction model to obtain a classification prediction result; and constructing an evaluation index to evaluate the performance of the method.