Design gene sequences to optimize expression using deep learning
usage: motif2json [-h] [--input INPUT] [--output OUTPUT]
This program takes the results from the motif finding and parses that file into a json file
-h, --help show this help message and exit
--input INPUT input motif generated by FIMO
--output OUTPUT output json file
merge the features into one big file
mpath=/research/gustavo1/CodonUsage
deepGene mergeFeatures --features $mpath/features/TFBs/HOCOMOCOv10.json $mpath/features/TFBs/miRNA.json $mpath/features/TFBs/RNA_Ray2013.json $mpath/features/TFBs/TFBSshape.json $mpath/features/TFBs/CIS-BP.json $mpath/features/TFBs/JASPAR_CORE_2016.json /research/shabbir5/AAV_project/output.json --genes $mpath/model/v2/labels.txt --output $mpath/model/v2/dataset.hdf5
Run the training
deepGene train --dataset $path/model/v2/dataset.hdf5 --model $path/model/v2/model.hdf5 --test 0.3 --epochs 180 --batch_size 64