/deepGene

gene expression optimization

Primary LanguagePython

DEEPGene

Design gene sequences to optimize expression using deep learning

Motif file from MEME to Json file

usage: motif2json [-h] [--input INPUT] [--output OUTPUT]

This program takes the results from the motif finding and parses that file into a json file

optional arguments:

-h, --help       show this help message and exit
--input INPUT    input motif generated by FIMO
--output OUTPUT  output json file

Usage

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