Extracts features using motif convolution

python -m pip install git+https://github.com/boooooogey/motif-Convo.git

To run it:

motifConvolve tfextract --meme meme_file --bed bed_file --genome genome.fa --mode max/average --out matrix_out_file

• meme_file is a MEME motif file.
• bed_file is a bed file that contains ATAC peaks.
• genome.fa is the FASTA file for the genome.
• After it calculates convolutions of motifs and peaks, it pools them either with max or average. You can specify this with --mode. The default is max.
• Once the execution is finished, it writes the resulting matrix to matrix_out_file . Rows of the matrix correspond to peaks, and columns correspond to motifs. There are 4 additional columns: GC_ratio, GC_pattern, CG_pattern, and Masked_ratio. GC_ratio is the ratio of the total number of G and C to the length of a peak. GC_pattern is the ratio of the GC pattern and CG_pattern is the ratio of the CG pattern within a peak. Masked_ratio is the ratio of the repressed sites to the length.
• You can change the batch size with --batch. It may affect the execution time. The default is 128.
• There is --window flag. Using it, you can change the window size around peaks. The default is 240.