main_single_prediction.py

from src import fastqc_extract, data_preparation, feature_engineering, ml_model
import sys

def main(fastq_file, model_dir, exports_dir, organism):
    if organism not in ['Ecoli', 'Efcm', 'Sau']: #replace with dynamic list of available models
        organism = 'complete_data'
 
    ngs_reads = fastqc_extract.import_all_reads(fastq_file, exports_dir, force_reimport=False, include_metadata=False, single_file=True)
    filenames = ngs_reads.index
    ngs_reads = data_preparation.prepare_fastqc_data(ngs_reads)
    ngs_reads = feature_engineering.apply_feature_engineering(ngs_reads, exports_dir, force_reimport=True)

    pred = ml_model.predict_evaluation(ngs_reads, model_dir+'/model_rf_'+organism+'.pkl')

    with open(exports_dir+'/predictions.txt', 'w') as f:
        for id, prediction in enumerate(pred):
            f.write(filenames[id]+': ')
            if prediction == 0:
                print('ugly')
                f.write('ugly\n')
            elif prediction == 1:
                print('good')
                f.write('good\n')

if __name__ == "__main__":
    fastq_file =  sys.argv[1] #'test_data/all/180525-18-3598-UW18720_S3_L001_R1_001.fastq.gz'
    model_dir = sys.argv[2] #'models'
    exports_dir = sys.argv[3] #'exports_evaluation_single'
    organism = sys.argv[4] #complete_data
    main(fastq_file, model_dir, exports_dir, organism)