Added script that takes in fastq files and converts them to fasta files with a filename that corresponds to the original fast5 read

Fixed bug where not providing --uncenter-kmer would cause snakemake to be confused (forgot this file)

Implemented predict function, can print confusion matrix and parse learning_rate and threshold keywords

# Conflicts:
#	nns/Rnn_keras.py

Fixed bug where too similar (SW) sequences slipped through initialisation and commented out saving figures from main script