Fig. 1
From: SeqEnhDL: sequence-based classification of cell type-specific enhancers using deep learning models
The SeqEnhDL approach. a Flowchart of the general SeqEnhDL procedure. A more detailed flowchart is available in the Additional file 1: Figure S1. b An intuitive example of the positional k-mer fold changes for sequence representation. The enhancer sequence at chr2: 182,807,955–182,808,154, with 5 bp flanking regions, is displayed. The example shows how to generate features for the 13th position (nucleotide “A”) among the 200 bp enhancer region. 5, 7, 9, and 11-mer centerred at the nucleotide “A” is extracted. Then, these k-mers are searched against dictionaries for their fold changes. Finally, the features at the 13th position are represented by the fold changes of its 5, 7, 9, and 11-mers