feizhihui / Protein-Methylation-Prediction

Sequence classification about protein methylation prediction

Protein Methylation Prediction

This repo include data normalization, Bi-directional RNN implemented by tensorflow-1.1.0.
And it is a sequence classification task about protein methylation prediction.

Max-min normalization

	min_v = np.min(self.train_prop1_data)
	max_v = np.max(self.train_prop1_data)
	self.train_prop1_data = (self.train_prop1_data - min_v) / (max_v - min_v)
	
	with open('../cache/maxmin_prob.txt', 'w') as file:
		file.write(str(min_v) + " " + str(max_v) + "\n")
	

Z-score normalization

	mu = np.mean(self.train_prop1_data)
	sigma = np.std(self.train_prop1_data)
	self.train_prop1_data = (self.train_prop1_data - mu) / sigma
	
	with open('../cache/zscore_prob.txt', 'w') as file:
		file.write(str(mu) + " " + str(sigma) + "\n")